Update README.md
Browse files
README.md
CHANGED
|
@@ -2,13 +2,14 @@
|
|
| 2 |
license: cc-by-nc-nd-3.0
|
| 3 |
quantized_by: bartowski
|
| 4 |
pipeline_tag: text-generation
|
|
|
|
| 5 |
---
|
| 6 |
|
| 7 |
-
## Llamacpp imatrix Quantizations of
|
| 8 |
|
| 9 |
Using <a href="https://github.com/ggerganov/llama.cpp/">llama.cpp</a> release <a href="https://github.com/ggerganov/llama.cpp/releases/tag/b2854">b2854</a> for quantization.
|
| 10 |
|
| 11 |
-
Original model: https://huggingface.co/
|
| 12 |
|
| 13 |
All quants made using imatrix option with dataset from [here](https://gist.github.com/bartowski1182/b6ac44691e994344625687afe3263b3a)
|
| 14 |
|
|
@@ -27,28 +28,28 @@ All quants made using imatrix option with dataset from [here](https://gist.githu
|
|
| 27 |
|
| 28 |
| Filename | Quant type | File Size | Description |
|
| 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 |
## Downloading using huggingface-cli
|
| 54 |
|
|
@@ -61,16 +62,16 @@ pip install -U "huggingface_hub[cli]"
|
|
| 61 |
Then, you can target the specific file you want:
|
| 62 |
|
| 63 |
```
|
| 64 |
-
huggingface-cli download bartowski/
|
| 65 |
```
|
| 66 |
|
| 67 |
If the model is bigger than 50GB, it will have been split into multiple files. In order to download them all to a local folder, run:
|
| 68 |
|
| 69 |
```
|
| 70 |
-
huggingface-cli download bartowski/
|
| 71 |
```
|
| 72 |
|
| 73 |
-
You can either specify a new local-dir (
|
| 74 |
|
| 75 |
## Which file should I choose?
|
| 76 |
|
|
|
|
| 2 |
license: cc-by-nc-nd-3.0
|
| 3 |
quantized_by: bartowski
|
| 4 |
pipeline_tag: text-generation
|
| 5 |
+
base_model: RLHFlow/LLaMA3-iterative-DPO-final
|
| 6 |
---
|
| 7 |
|
| 8 |
+
## Llamacpp imatrix Quantizations of LLaMA3-iterative-DPO-final
|
| 9 |
|
| 10 |
Using <a href="https://github.com/ggerganov/llama.cpp/">llama.cpp</a> release <a href="https://github.com/ggerganov/llama.cpp/releases/tag/b2854">b2854</a> for quantization.
|
| 11 |
|
| 12 |
+
Original model: https://huggingface.co/RLHFlow/LLaMA3-iterative-DPO-final
|
| 13 |
|
| 14 |
All quants made using imatrix option with dataset from [here](https://gist.github.com/bartowski1182/b6ac44691e994344625687afe3263b3a)
|
| 15 |
|
|
|
|
| 28 |
|
| 29 |
| Filename | Quant type | File Size | Description |
|
| 30 |
| -------- | ---------- | --------- | ----------- |
|
| 31 |
+
| [LLaMA3-iterative-DPO-final-Q8_0.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-Q8_0.gguf) | Q8_0 | 8.54GB | Extremely high quality, generally unneeded but max available quant. |
|
| 32 |
+
| [LLaMA3-iterative-DPO-final-Q6_K.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-Q6_K.gguf) | Q6_K | 6.59GB | Very high quality, near perfect, *recommended*. |
|
| 33 |
+
| [LLaMA3-iterative-DPO-final-Q5_K_M.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-Q5_K_M.gguf) | Q5_K_M | 5.73GB | High quality, *recommended*. |
|
| 34 |
+
| [LLaMA3-iterative-DPO-final-Q5_K_S.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-Q5_K_S.gguf) | Q5_K_S | 5.59GB | High quality, *recommended*. |
|
| 35 |
+
| [LLaMA3-iterative-DPO-final-Q4_K_M.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-Q4_K_M.gguf) | Q4_K_M | 4.92GB | Good quality, uses about 4.83 bits per weight, *recommended*. |
|
| 36 |
+
| [LLaMA3-iterative-DPO-final-Q4_K_S.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-Q4_K_S.gguf) | Q4_K_S | 4.69GB | Slightly lower quality with more space savings, *recommended*. |
|
| 37 |
+
| [LLaMA3-iterative-DPO-final-IQ4_NL.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-IQ4_NL.gguf) | IQ4_NL | 4.67GB | Decent quality, slightly smaller than Q4_K_S with similar performance *recommended*. |
|
| 38 |
+
| [LLaMA3-iterative-DPO-final-IQ4_XS.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-IQ4_XS.gguf) | IQ4_XS | 4.44GB | Decent quality, smaller than Q4_K_S with similar performance, *recommended*. |
|
| 39 |
+
| [LLaMA3-iterative-DPO-final-Q3_K_L.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-Q3_K_L.gguf) | Q3_K_L | 4.32GB | Lower quality but usable, good for low RAM availability. |
|
| 40 |
+
| [LLaMA3-iterative-DPO-final-Q3_K_M.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-Q3_K_M.gguf) | Q3_K_M | 4.01GB | Even lower quality. |
|
| 41 |
+
| [LLaMA3-iterative-DPO-final-IQ3_M.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-IQ3_M.gguf) | IQ3_M | 3.78GB | Medium-low quality, new method with decent performance comparable to Q3_K_M. |
|
| 42 |
+
| [LLaMA3-iterative-DPO-final-IQ3_S.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-IQ3_S.gguf) | IQ3_S | 3.68GB | Lower quality, new method with decent performance, recommended over Q3_K_S quant, same size with better performance. |
|
| 43 |
+
| [LLaMA3-iterative-DPO-final-Q3_K_S.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-Q3_K_S.gguf) | Q3_K_S | 3.66GB | Low quality, not recommended. |
|
| 44 |
+
| [LLaMA3-iterative-DPO-final-IQ3_XS.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-IQ3_XS.gguf) | IQ3_XS | 3.51GB | Lower quality, new method with decent performance, slightly better than Q3_K_S. |
|
| 45 |
+
| [LLaMA3-iterative-DPO-final-IQ3_XXS.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-IQ3_XXS.gguf) | IQ3_XXS | 3.27GB | Lower quality, new method with decent performance, comparable to Q3 quants. |
|
| 46 |
+
| [LLaMA3-iterative-DPO-final-Q2_K.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-Q2_K.gguf) | Q2_K | 3.17GB | Very low quality but surprisingly usable. |
|
| 47 |
+
| [LLaMA3-iterative-DPO-final-IQ2_M.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-IQ2_M.gguf) | IQ2_M | 2.94GB | Very low quality, uses SOTA techniques to also be surprisingly usable. |
|
| 48 |
+
| [LLaMA3-iterative-DPO-final-IQ2_S.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-IQ2_S.gguf) | IQ2_S | 2.75GB | Very low quality, uses SOTA techniques to be usable. |
|
| 49 |
+
| [LLaMA3-iterative-DPO-final-IQ2_XS.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-IQ2_XS.gguf) | IQ2_XS | 2.60GB | Very low quality, uses SOTA techniques to be usable. |
|
| 50 |
+
| [LLaMA3-iterative-DPO-final-IQ2_XXS.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-IQ2_XXS.gguf) | IQ2_XXS | 2.39GB | Lower quality, uses SOTA techniques to be usable. |
|
| 51 |
+
| [LLaMA3-iterative-DPO-final-IQ1_M.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-IQ1_M.gguf) | IQ1_M | 2.16GB | Extremely low quality, *not* recommended. |
|
| 52 |
+
| [LLaMA3-iterative-DPO-final-IQ1_S.gguf](https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF/blob/main/LLaMA3-iterative-DPO-final-IQ1_S.gguf) | IQ1_S | 2.01GB | Extremely low quality, *not* recommended. |
|
| 53 |
|
| 54 |
## Downloading using huggingface-cli
|
| 55 |
|
|
|
|
| 62 |
Then, you can target the specific file you want:
|
| 63 |
|
| 64 |
```
|
| 65 |
+
huggingface-cli download bartowski/LLaMA3-iterative-DPO-final-GGUF --include "LLaMA3-iterative-DPO-final-Q4_K_M.gguf" --local-dir ./ --local-dir-use-symlinks False
|
| 66 |
```
|
| 67 |
|
| 68 |
If the model is bigger than 50GB, it will have been split into multiple files. In order to download them all to a local folder, run:
|
| 69 |
|
| 70 |
```
|
| 71 |
+
huggingface-cli download bartowski/LLaMA3-iterative-DPO-final-GGUF --include "LLaMA3-iterative-DPO-final-Q8_0.gguf/*" --local-dir LLaMA3-iterative-DPO-final-Q8_0 --local-dir-use-symlinks False
|
| 72 |
```
|
| 73 |
|
| 74 |
+
You can either specify a new local-dir (LLaMA3-iterative-DPO-final-Q8_0) or download them all in place (./)
|
| 75 |
|
| 76 |
## Which file should I choose?
|
| 77 |
|