Llamacpp Quantizations of Magic-Dolphin-7b
Using llama.cpp commit fa97464 for quantization.
Original model: https://huggingface.co/InferenceIllusionist/Magic-Dolphin-7b
Download a file (not the whole branch) from below:
Filename | Quant type | File Size | Description |
---|---|---|---|
Magic-Dolphin-7b-Q8_0.gguf | Q8_0 | 7.69GB | Extremely high quality, generally unneeded but max available quant. |
Magic-Dolphin-7b-Q6_K.gguf | Q6_K | 5.94GB | Very high quality, near perfect, recommended. |
Magic-Dolphin-7b-Q5_K_M.gguf | Q5_K_M | 5.13GB | High quality, very usable. |
Magic-Dolphin-7b-Q5_K_S.gguf | Q5_K_S | 4.99GB | High quality, very usable. |
Magic-Dolphin-7b-Q5_0.gguf | Q5_0 | 4.99GB | High quality, older format, generally not recommended. |
Magic-Dolphin-7b-Q4_K_M.gguf | Q4_K_M | 4.36GB | Good quality, similar to 4.25 bpw. |
Magic-Dolphin-7b-Q4_K_S.gguf | Q4_K_S | 4.14GB | Slightly lower quality with small space savings. |
Magic-Dolphin-7b-Q4_0.gguf | Q4_0 | 4.10GB | Decent quality, older format, generally not recommended. |
Magic-Dolphin-7b-Q3_K_L.gguf | Q3_K_L | 3.82GB | Lower quality but usable, good for low RAM availability. |
Magic-Dolphin-7b-Q3_K_M.gguf | Q3_K_M | 3.51GB | Even lower quality. |
Magic-Dolphin-7b-Q3_K_S.gguf | Q3_K_S | 3.16GB | Low quality, not recommended. |
Magic-Dolphin-7b-Q2_K.gguf | Q2_K | 2.71GB | Extremely low quality, not recommended. |
Want to support my work? Visit my ko-fi page here: https://ko-fi.com/bartowski
- Downloads last month
- 1,259
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Model tree for bartowski/Magic-Dolphin-7b-GGUF
Merge model
this model
Evaluation results
- normalized accuracy on AI2 Reasoning Challenge (25-Shot)test set Open LLM Leaderboard65.780
- normalized accuracy on HellaSwag (10-Shot)validation set Open LLM Leaderboard85.610
- accuracy on MMLU (5-Shot)test set Open LLM Leaderboard64.640
- mc2 on TruthfulQA (0-shot)validation set Open LLM Leaderboard58.010
- accuracy on Winogrande (5-shot)validation set Open LLM Leaderboard79.640
- accuracy on GSM8k (5-shot)test set Open LLM Leaderboard51.180