This model is a fine-tuned version of Wav2Vec2-BERT 2.0 for Kinyarwanda automatic speech recognition (ASR). It was trained on the 1000 hours dataset from the Kinyarwanda ASR hackthon on Kaggle (Track B), dataset covering Health, Government, Finance, Education, and Agriculture domains. The model is robust and the in-domain WER is below 8.4%. - **Developed by:** Badr al-Absi - **Model type:** Speech Recognition (ASR) - **Language:** Kinyarwanda (rw) - **License:** CC-BY-4.0 - **Finetuned from:** facebook/w2v-bert-2.0 - **Repository:** https://huggingface.co/badrex/w2v-bert-2.0-kinyarwanda-asr - **Dataset:** Kinyarwanda ASR Track B
翻訳待ち・原文を表示
入力 音声→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
入力 テキスト→ 出力 テキスト
**The whole 35B model in a single 9.96 GB file.** 2.30 bits per weight, and it retains 94.1% of the full-precision model's measured behaviour. BTL-4 is a mixture of experts with roughly 2.1B active parameters per token, so it costs a large model's memory and a small model's compute. Compact is the edition that runs on hardware you already own — one file, one command, a running agent. No base download, no reconstruction. Loads in llama.cpp, Ollama and LM Studio.
翻訳待ち・原文を表示
入力 テキスト→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
入力 テキスト · 画像→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
入力 テキスト→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
入力 テキスト→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
入力 テキスト · 画像→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
入力 テキスト · 画像→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
入力 テキスト→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
入力 テキスト · 画像→ 出力 テキスト
> [!WARNING] > **Experimental low-bit release.** The routed experts, which hold most of the > weights, are stored at 1.625 or 2.625 bits per weight depending on the layer > (about 2.1 on average), far below common 4-bit GGUF quants. You may see > failures the source model rarely shows: reasoning that loops or never > reaches an answer, text that breaks mid-character (shown as `�`) or switches > language mid-word, and weaker answers in general. This is exploratory > quantization research, not a drop-in replacement for the source model. > > **The files can change.** This line of conversions is still being researched > (as of 2026-10-21, weighted Lloyd fitting of the codebooks is being tested), > so any file here may be replaced by a new revision. Replacements so far: two > since the first release on 2026-10-17. The 8-bit residual stage that > protects sensitive routed layers was added to layers 3–16 and then widened > to 19 layers (2026-10-18). The earlier files remain at the tags `q4k-v3` and > `r13p8-l3-16`. To keep a fixed build, download a pinned commit: > `hf download Baekpica/GLM-5.3-Flash-Uncensored-VQ-GGUF --revision `. > > **Problems or odd output?** Feel free to post them in the > Community tab, > even as rough notes. The prompt, sampling settings and runtime commit help > most. > **Status, 2026-10-18:** the files are the 19-layer R13+8 revision: the > residual stage now covers routed layers 3–17, 22, 27, 41 and 44 (before: > 3–16). Loading needs a ds4-dfm-rs build with GVQ1 container v2 support > (`main` since PR #96, > merge `34c46d18`); the runtime binds the residual stage on any routed layer, > so no runtime change is needed. Speed and memory of this revision on DGX > Spark are not measured yet. Previous revisions are kept at the tags > `r13p8-l3-16` > (R13+8 on layers 3–16) and > `q4k-v3` > (all routed layers R13). Experimental custom GVQ1 / GGUF v3 conversion of orcarouter/GLM-5.3-Flash-Uncensored-FP8 (`150bd5e3a957db747db1224cc4f8891bcfa7a7f0`), the abliterated (refusal-removed) build of Z.AI GLM-5.3-Flash in the official block-FP8 layout, for DGX Spark (GB10) with ds4-dfm-rs. Standard llama.cpp/Ollama/Transformers loading is not claimed.
翻訳待ち・原文を表示
入力 テキスト · 画像→ 出力 テキスト
> [!WARNING] > **Experimental low-bit release.** The routed experts, which hold most of the > weights, are stored at 1.625 or 2.625 bits per weight depending on the layer > (about 2.1 on average), far below common 4-bit GGUF quants. You may see > failures the source model rarely shows: reasoning that loops or never > reaches an answer, text that breaks mid-character (shown as `�`) or switches > language mid-word, and weaker answers in general. This is exploratory > quantization research, not a drop-in replacement for the source model. > > **The files can change.** This line of conversions is still being researched > (as of 2026-10-21, weighted Lloyd fitting of the codebooks is being tested), > so any file here may be replaced by a new revision. Replacements so far: > three since the first release on 2026-10-15. The non-expert weights were > requantized (Q4_K v3, 2026-10-16), then the 8-bit residual stage that > protects sensitive routed layers was added to layers 3–16 and widened to 19 > layers (2026-10-18). The earlier files remain at the tags `vq-original`, > `q4k-v3` and `r13p8-l3-16`. To keep a fixed build, download a pinned commit: > `hf download Baekpica/GLM-5.3-Flash-VQ-GGUF --revision `. > > **Problems or odd output?** Feel free to post them in the > Community tab, > even as rough notes. The prompt, sampling settings and runtime commit help > most. > **Status, 2026-10-18:** the files are the 19-layer R13+8 revision: the > residual stage now covers routed layers 3–17, 22, 27, 41 and 44 (before: > 3–16). Loading needs a ds4-dfm-rs build with GVQ1 container v2 support > (`main` since PR #96, > merge `34c46d18`); the runtime binds the residual stage on any routed layer, > so no runtime change is needed. Speed and memory of this revision on DGX > Spark are not measured yet. Previous revisions are kept at the tags > `r13p8-l3-16` > (R13+8 on layers 3–16) and > `q4k-v3` > (all routed layers R13). Experimental custom GVQ1 / GGUF v3 conversion of the official Z.AI GLM-5.3-Flash (`91b69215804f2d2036a8867f6ce44848e278dcda`) for DGX Spark (GB10), run by ds4-dfm-rs. Standard llama.cpp/Ollama/Transformers loading is not claimed.
翻訳待ち・原文を表示
入力 テキスト · 画像→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
入力 テキスト→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
分類準備中
入力 未確認→ 出力 未確認
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
入力 テキスト · 画像→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
入力 テキスト→ 出力 テキスト
> **Mixed weights with original-representation calibration.** The four-shard language model passed its 508-tensor structural and quantization audit. > > **Role-aware precision:** expert gate/up **IQ2_XXS**, expert down **IQ2_XS**, shared dense paths and embedded MTP **Q8_0**, media **BF16**, and numerical controls **F32**. > > **Measured GGUF storage: 93.092 GB / 86.699 GiB**, including main model, multimodal projector and separate DFlash weights. Auxiliary files and runtime memory are additional. > > Text, image, audio, video and audiovisual assets are retained. **ds4-dfm-rs on one DGX Spark GB10:** the bounded serving checks below cover two 256K text banks, serial media, live partial reuse and restart disk KV with MTP off. Prior 512K serial text and 256K serial media/DFlash checks used different configurations. The checks do not establish throughput parity or general quality retention. The 256K mixed setup uses `--max-seqs 2 --prefix-reuse partial --kv-disk-dir ... --mtp-mode off`, `DS4_MIMO2_PREFILL_CHUNK=2048`, and a VMM owner with `--no-repack-q8-aligned`. Aligned IQ2 expert weights stay enabled; disabling the optional Q8 repack saves 6.26 GiB. The two text banks hold separate KV and serialize forward passes through shared scratch. Images, video and audio use the serial lane. These are bounded functional checks, not a full-length 256K mixed request, multimodal quality, cache reuse at 1M, or throughput qualification. The serving report records the exact limits. Embedded MTP without DFlash and the external five-layer DFlash with serial media were checked separately before this setup.
翻訳待ち・原文を表示
入力 テキスト · 画像→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
入力 テキスト · 画像→ 出力 テキスト
> [!WARNING] > **Problems or odd output?** Post them in the > Community tab, even as > rough notes. The prompt, sampling settings and runtime commit help most. **MiniMax's 428B native multimodal MoE with its full 1,048,576-token context, its sparse attention, its vision/video tower and its DSpark draft head, sized for one 128 GB DGX Spark.** The 236 GB checkpoint (NVFP4 experts, MXFP8 attention) becomes **115.5 GB** of GGUF parts, of which **111.3 GB (103.7 GiB) stay resident**; the BF16 token table and the vision tower are read from SSD, and the attention K/V of a long session lives on SSD, addressed block by block by the model's own MSA indexer. > **Runtime support, 2026-10-24:** runs on one DGX Spark with the ds4-dfm-rs `minimax_m3_vq` family > (merged PR #110): text, images and video, the DSpark draft head and the native context. > **Performance, 2026-10-29:** matched 2K–64K Prefill **1,001.85 → 1,020.29 tok/s (+1.84%)**; > the last 2K at 64K reaches **990.08 tok/s (+2.87%)**, using the runtime changes in > PR #126 (conditions and Decode limits below). > The historical 1,048,319-token prompt prefilled at 580 tok/s and > decoded. Converted 2026-10-23: 61 independent GGUF parts (115,526,325,376 bytes), every payload > hash in `tensor-manifest.json`, all 89,729 source tensors accounted for; codebooks fitted with > calibration statistics from 4.18M tokens (text, images and video) run through the original > checkpoint. **Measured against the original, teacher-forced on held-out chats: KL 0.150 and the > same next token 88.2% of the time on assistant turns** (the original batched differently: KL > 0.003). The memory and kernel budgets below are the design computed from exact byte layouts.
翻訳待ち・原文を表示
入力 テキスト · 画像→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
入力 テキスト→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
入力 テキスト · 画像→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
入力 テキスト · 画像→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
入力 テキスト→ 出力 テキスト
基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
基本情報を収録・詳細は順次追加
入力 テキスト · 画像→ 出力 テキスト
