Views
No views yet
| Name | Quant method | Size |
|---|---|---|
| GritLM-7B.Q2_K.gguf | Q2_K | 2.53GB |
| GritLM-7B.IQ3_XS.gguf | IQ3_XS | 2.81GB |
| GritLM-7B.IQ3_S.gguf | IQ3_S | 2.96GB |
| GritLM-7B.Q3_K_S.gguf | Q3_K_S | 2.95GB |
| GritLM-7B.IQ3_M.gguf | IQ3_M | 3.06GB |
| GritLM-7B.Q3_K.gguf | Q3_K | 3.28GB |
| GritLM-7B.Q3_K_M.gguf | Q3_K_M | 3.28GB |
| GritLM-7B.Q3_K_L.gguf | Q3_K_L | 3.56GB |
| GritLM-7B.IQ4_XS.gguf | IQ4_XS | 3.67GB |
| GritLM-7B.Q4_0.gguf | Q4_0 | 3.83GB |
| GritLM-7B.IQ4_NL.gguf | IQ4_NL | 3.87GB |
| GritLM-7B.Q4_K_S.gguf | Q4_K_S | 3.86GB |
| GritLM-7B.Q4_K.gguf | Q4_K | 4.07GB |
| GritLM-7B.Q4_K_M.gguf | Q4_K_M | 4.07GB |
| GritLM-7B.Q4_1.gguf | Q4_1 | 4.24GB |
| GritLM-7B.Q5_0.gguf | Q5_0 | 4.65GB |
| GritLM-7B.Q5_K_S.gguf | Q5_K_S | 4.65GB |
| GritLM-7B.Q5_K.gguf | Q5_K | 4.78GB |
| GritLM-7B.Q5_K_M.gguf | Q5_K_M | 4.78GB |
| GritLM-7B.Q5_1.gguf | Q5_1 | 5.07GB |
| GritLM-7B.Q6_K.gguf | Q6_K | 5.53GB |
GritLM is a generative representational instruction tuned language model. It unifies text representation (embedding) and text generation into a single model achieving state-of-the-art performance on both types of tasks.
| Model | Description |
|---|---|
| GritLM 7B | Mistral 7B finetuned using GRIT |
| GritLM 8x7B | Mixtral 8x7B finetuned using GRIT |
1@misc{muennighoff2024generative,
2 title={Generative Representational Instruction Tuning},
3 author={Niklas Muennighoff and Hongjin Su and Liang Wang and Nan Yang and Furu Wei and Tao Yu and Amanpreet Singh and Douwe Kiela},
4 year={2024},
5 eprint={2402.09906},
6 archivePrefix={arXiv},
7 primaryClass={cs.CL}
8}