--- language: - en library_name: transformers license: llama2 quantized_by: mradermacher tags: - Xwin - WinterGoddess - frankenmerge - 120b --- ## About weighted/imatrix quants of https://huggingface.co/llmixer/Xwinter-120b static quants are available at https://huggingface.co/mradermacher/Xwinter-120b-GGUF ## Usage If you are unsure how to use GGUF files, refer to one of [TheBloke's READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for more details, including on how to concatenate multi-part files. ## Provided Quants (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants) | Link | Type | Size/GB | Notes | |:-----|:-----|--------:|:------| | [GGUF](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-IQ2_M.gguf) | i1-IQ2_M | 40.0 | | | [GGUF](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q2_K.gguf) | i1-Q2_K | 43.7 | IQ3_XXS probably better | | [GGUF](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 45.6 | fast, lower quality | | [GGUF](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-IQ3_XS.gguf) | i1-IQ3_XS | 48.4 | | | [PART 1](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q3_K_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q3_K_S.gguf.part2of2) | i1-Q3_K_S | 51.1 | IQ3_XS probably better | | [PART 1](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q3_K_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q3_K_M.gguf.part2of2) | i1-Q3_K_M | 57.0 | IQ3_S probably better | | [PART 1](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q3_K_L.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q3_K_L.gguf.part2of2) | i1-Q3_K_L | 62.1 | IQ3_M probably better | | [PART 1](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q4_K_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q4_K_S.gguf.part2of2) | i1-Q4_K_S | 67.2 | optimal size/speed/quality | | [PART 1](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q4_K_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q4_K_M.gguf.part2of2) | i1-Q4_K_M | 71.0 | fast, medium quality | | [PART 1](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q5_K_S.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q5_K_S.gguf.part2of2) | i1-Q5_K_S | 81.4 | | | [PART 1](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q5_K_M.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q5_K_M.gguf.part2of2) | i1-Q5_K_M | 83.6 | | | [PART 1](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q6_K.gguf.part1of3) [PART 2](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q6_K.gguf.part2of3) [PART 3](https://huggingface.co/mradermacher/Xwinter-120b-i1-GGUF/resolve/main/Xwinter-120b.i1-Q6_K.gguf.part3of3) | i1-Q6_K | 97.0 | practically like static Q6_K | Here is a handy graph by ikawrakow comparing some lower-quality quant types (lower is better): ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png) And here are Artefact2's thoughts on the matter: https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9 ## Thanks I thank my company, [nethype GmbH](https://www.nethype.de/), for letting me use its servers and providing upgrades to my workstation to enable this work in my free time.