--- title: Index-Echo-S2ST-2B-GGUF canonical_url: "https://www.modelscope.ai/models/IndexTeam/Index-Echo-S2ST-2B-GGUF" md_url: "https://www.modelscope.ai/models/IndexTeam/Index-Echo-S2ST-2B-GGUF.md" repository: IndexTeam/Index-Echo-S2ST-2B-GGUF last_updated: 2026-10-16 license: apache-2.0 base_model: - IndexTeam/Index-Echo-S2ST-2B base_model_relation: quantized library_name: - gguf - pytorch frameworks: - pytorch downloads: 207 stars: 0 tags: - gguf - translation - index --- # Index-Echo-S2ST-2B-GGUF > Index-Echo-S2ST-2B-GGUF - An open-source model by IndexTeam on ModelScope. Index-Echo-S2ST-2B-GGUF IndexTeam/Index-Echo-S2ST-2B-GGUF is a machine learning model on ModelScope. licensed under apache-2.0. derived from IndexTeam/Index-Echo-S2ST-2B. - **Repository**: IndexTeam/Index-Echo-S2ST-2B-GGUF - **License**: apache-2.0 - **Base model**: IndexTeam/Index-Echo-S2ST-2B - **Tags**: gguf, translation, index - **Downloads**: 207 - **Stars**: 0 - **Last updated**: 2026-10-16 Source: https://www.modelscope.ai/models/IndexTeam/Index-Echo-S2ST-2B-GGUF --- # Index-Echo-S2ST-2B-GGUF Official GGUF conversion of the **text LLM backbone** (`stlm_llm/`) of [IndexTeam/Index-Echo-S2ST-2B](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B) (**v4.1**), part of the **Index-Echo** speech-to-speech translation (S2ST) model family by bilibili. The v4.1 package updates the multilingual translator and extends the release to **twelve directions**: Chinese, Japanese, or English speech input, dubbed into English, Spanish, Japanese, Korean, or Chinese. > **Note: LLM backbone only.** The audio tower, connector and speech-synthesis stack are **not** included - llama.cpp cannot run the full speech pipeline. For full speech-to-speech translation (S2ST) inference, use the original checkpoint or the FP8 build in [Index-Echo-S2ST-2B-FP8](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B-FP8), which ships the complete pipeline with an FP8-quantized LLM. Converted with [llama.cpp](https://github.com/ggml-org/llama.cpp) (master, 2026-10-13); static post-training quantization, all bit-widths in this single repository. ## Provided files (sorted by size) | File | Size | Notes | |:-----|-----:|:------| | [mmproj-Q8_0](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B-GGUF/resolve/main/Index-Echo-S2ST-2B.mmproj-Q8_0.gguf) | 0.36 GB | multi-modal projector supplement | | [mmproj-f16](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B-GGUF/resolve/main/Index-Echo-S2ST-2B.mmproj-f16.gguf) | 0.67 GB | multi-modal projector supplement | | [Q2_K](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B-GGUF/resolve/main/Index-Echo-S2ST-2B.Q2_K.gguf) | 0.97 GB | 2-bit, significant quality loss | | [Q3_K_S](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B-GGUF/resolve/main/Index-Echo-S2ST-2B.Q3_K_S.gguf) | 1.02 GB | 3-bit, noticeable quality loss | | [Q3_K_M](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B-GGUF/resolve/main/Index-Echo-S2ST-2B.Q3_K_M.gguf) | 1.10 GB | 3-bit, noticeable quality loss | | [Q3_K_L](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B-GGUF/resolve/main/Index-Echo-S2ST-2B.Q3_K_L.gguf) | 1.16 GB | 3-bit, noticeable quality loss | | [IQ4_XS](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B-GGUF/resolve/main/Index-Echo-S2ST-2B.IQ4_XS.gguf) | 1.20 GB | 4-bit I-quant, small | | [Q4_K_S](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B-GGUF/resolve/main/Index-Echo-S2ST-2B.Q4_K_S.gguf) | 1.21 GB | 4-bit, slightly smaller/faster | | [Q4_K_M](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B-GGUF/resolve/main/Index-Echo-S2ST-2B.Q4_K_M.gguf) | 1.27 GB | 4-bit, good balance, recommended | | [Q5_K_S](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B-GGUF/resolve/main/Index-Echo-S2ST-2B.Q5_K_S.gguf) | 1.37 GB | 5-bit, low quality loss | | [Q5_K_M](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B-GGUF/resolve/main/Index-Echo-S2ST-2B.Q5_K_M.gguf) | 1.41 GB | 5-bit, low quality loss | | [Q6_K](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B-GGUF/resolve/main/Index-Echo-S2ST-2B.Q6_K.gguf) | 1.56 GB | 6-bit, very low quality loss | | [Q8_0](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B-GGUF/resolve/main/Index-Echo-S2ST-2B.Q8_0.gguf) | 2.01 GB | 8-bit, near-lossless | | [f16](https://huggingface.co/IndexTeam/Index-Echo-S2ST-2B-GGUF/resolve/main/Index-Echo-S2ST-2B.f16.gguf) | 3.78 GB | 16-bit, lossless conversion baseline | The `mmproj` files are the multi-modal projector for the vision tower; they are only needed for image-input use with `llama-mtmd-cli`. For text-only translation, the plain GGUF files suffice. (The v4.1 package no longer bundles this tower inside `stlm_llm/`; the `mmproj` files are carried over unchanged from the initial release.) ## Usage ```bash # Run the recommended Q4_K_M directly: llama serve -hf IndexTeam/Index-Echo-S2ST-2B-GGUF:Q4_K_M llama cli -hf IndexTeam/Index-Echo-S2ST-2B-GGUF:Q4_K_M ``` ## Consistency validation Before release, each quantization tier was validated against the F16 conversion: per-token KL divergence / RMS delta-p via `llama-perplexity` (CPU build of llama.cpp), plus greedy-generation spot checks against the original BF16 weights (transformers reference, CPU). | Quant | PPL (fixed corpus) | Mean KLD | RMS Δp | Same top-1 | |:------|-------------------:|---------:|-------:|-----------:| | IQ4_XS | 4.9447 | 0.071674 | 7.86% | 91.13% | | Q2_K | 10.6969 | 1.090653 | 33.49% | 63.01% | | Q3_K_L | 5.4337 | 0.217771 | 13.94% | 84.47% | | Q3_K_M | 5.4265 | 0.244748 | 14.17% | 83.30% | | Q3_K_S | 6.0653 | 0.349583 | 17.79% | 80.43% | | Q4_K_M | 4.9306 | 0.070487 | 7.82% | 90.93% | | Q4_K_S | 4.8496 | 0.074472 | 7.77% | 90.74% | | Q5_K_M | 4.7035 | 0.020143 | 4.31% | 94.98% | | Q5_K_S | 4.6823 | 0.021831 | 4.37% | 94.85% | | Q6_K | 4.7519 | 0.008402 | 2.67% | 96.48% | | Q8_0 | 4.7325 | 0.001318 | 1.20% | 99.02% | | f16 | 4.7059 | (reference) | - | - | *Converted and published by the Index team, 2026-10-13.*