--- base_model: - cognitivecomputations/dolphin-2.2-70b - cognitivecomputations/dolphin-2.2-70b - cognitivecomputations/dolphin-2.2-70b - cognitivecomputations/dolphin-2.2-70b - cognitivecomputations/dolphin-2.2-70b - cognitivecomputations/dolphin-2.2-70b - cognitivecomputations/dolphin-2.2-70b datasets: - ehartford/dolphin - jondurbin/airoboros-2.2.1 - ehartford/samantha-data - ehartford/WizardLM_evol_instruct_V2_196k_unfiltered_merged_split language: - en library_name: transformers license: llama2 quantized_by: mradermacher --- ## About weighted/imatrix quants of https://huggingface.co/cognitivecomputations/MegaDolphin-120b ## 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/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-IQ1_S.gguf) | i1-IQ1_S | 25.6 | | | [GGUF](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 32.1 | | | [GGUF](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-IQ2_XS.gguf) | i1-IQ2_XS | 35.7 | | | [GGUF](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-Q2_K.gguf) | i1-Q2_K | 44.5 | IQ3_XXS probably better | | [GGUF](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 47.2 | fast, lower quality | | [PART 1](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-Q3_K_XS.gguf.split-aa) [PART 2](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-Q3_K_XS.gguf.split-ab) | i1-Q3_K_XS | 49.2 | | | [PART 1](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-Q3_K_S.gguf.split-aa) [PART 2](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-Q3_K_S.gguf.split-ab) | i1-Q3_K_S | 52.1 | IQ3_XS probably better | | [PART 1](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-Q3_K_M.gguf.split-aa) [PART 2](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-Q3_K_M.gguf.split-ab) | i1-Q3_K_M | 58.1 | IQ3_S probably better | | [PART 1](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-Q3_K_L.gguf.split-aa) [PART 2](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-Q3_K_L.gguf.split-ab) | i1-Q3_K_L | 63.3 | IQ3_M probably better | | [PART 1](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-Q4_K_S.gguf.split-aa) [PART 2](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-Q4_K_S.gguf.split-ab) | i1-Q4_K_S | 68.6 | almost as good as Q4_K_M | | [PART 1](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-Q4_K_M.gguf.split-aa) [PART 2](https://huggingface.co/mradermacher/MegaDolphin-120b-i1-GGUF/resolve/main/MegaDolphin-120b.i1-Q4_K_M.gguf.split-ab) | i1-Q4_K_M | 72.5 | fast, medium quality | 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