File size: 5,228 Bytes
8d0e972 8e529d4 8d0e972 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 |
---
base_model: SultanR/SmolTulu-1.7b-RM
datasets:
- allenai/llama-3.1-tulu-3-8b-preference-mixture
language:
- en
library_name: transformers
license: apache-2.0
quantized_by: mradermacher
tags:
- Tulu3
- Smollm
- SLMs
- Small
- Huggingface
- Allenai
- Reward Model
- RLVR
- RM
- Reward
---
## About
<!-- ### quantize_version: 2 -->
<!-- ### output_tensor_quantised: 1 -->
<!-- ### convert_type: hf -->
<!-- ### vocab_type: -->
<!-- ### tags: nicoboss -->
weighted/imatrix quants of https://huggingface.co/SultanR/SmolTulu-1.7b-RM
<!-- provided-files -->
static quants are available at https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-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/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-IQ1_S.gguf) | i1-IQ1_S | 0.5 | for the desperate |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-IQ1_M.gguf) | i1-IQ1_M | 0.5 | mostly desperate |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 0.6 | |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-IQ2_XS.gguf) | i1-IQ2_XS | 0.6 | |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-IQ2_S.gguf) | i1-IQ2_S | 0.7 | |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-IQ2_M.gguf) | i1-IQ2_M | 0.7 | |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-Q2_K_S.gguf) | i1-Q2_K_S | 0.7 | very low quality |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-Q2_K.gguf) | i1-Q2_K | 0.8 | IQ3_XXS probably better |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 0.8 | lower quality |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-IQ3_XS.gguf) | i1-IQ3_XS | 0.8 | |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-IQ3_S.gguf) | i1-IQ3_S | 0.9 | beats Q3_K* |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-Q3_K_S.gguf) | i1-Q3_K_S | 0.9 | IQ3_XS probably better |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-IQ3_M.gguf) | i1-IQ3_M | 0.9 | |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-Q3_K_M.gguf) | i1-Q3_K_M | 1.0 | IQ3_S probably better |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-Q3_K_L.gguf) | i1-Q3_K_L | 1.0 | IQ3_M probably better |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-IQ4_XS.gguf) | i1-IQ4_XS | 1.0 | |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-Q4_0.gguf) | i1-Q4_0 | 1.1 | fast, low quality |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-Q4_K_S.gguf) | i1-Q4_K_S | 1.1 | optimal size/speed/quality |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-Q4_K_M.gguf) | i1-Q4_K_M | 1.2 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-Q5_K_S.gguf) | i1-Q5_K_S | 1.3 | |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-Q5_K_M.gguf) | i1-Q5_K_M | 1.3 | |
| [GGUF](https://huggingface.co/mradermacher/SmolTulu-1.7b-RM-i1-GGUF/resolve/main/SmolTulu-1.7b-RM.i1-Q6_K.gguf) | i1-Q6_K | 1.5 | practically like static Q6_K |
Here is a handy graph by ikawrakow comparing some lower-quality quant
types (lower is better):
And here are Artefact2's thoughts on the matter:
https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
## FAQ / Model Request
See https://huggingface.co/mradermacher/model_requests for some answers to
questions you might have and/or if you want some other model quantized.
## 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. Additional thanks to [@nicoboss](https://huggingface.co/nicoboss) for giving me access to his private supercomputer, enabling me to provide many more imatrix quants, at much higher quality, than I would otherwise be able to.
<!-- end -->
|