RichardErkhov
commited on
Commit
•
e4d656c
1
Parent(s):
7fba2a1
uploaded readme
Browse files
README.md
ADDED
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
Quantization made by Richard Erkhov.
|
2 |
+
|
3 |
+
[Github](https://github.com/RichardErkhov)
|
4 |
+
|
5 |
+
[Discord](https://discord.gg/pvy7H8DZMG)
|
6 |
+
|
7 |
+
[Request more models](https://github.com/RichardErkhov/quant_request)
|
8 |
+
|
9 |
+
|
10 |
+
Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B - GGUF
|
11 |
+
- Model creator: https://huggingface.co/yunconglong/
|
12 |
+
- Original model: https://huggingface.co/yunconglong/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B/
|
13 |
+
|
14 |
+
|
15 |
+
| Name | Quant method | Size |
|
16 |
+
| ---- | ---- | ---- |
|
17 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q2_K.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q2_K.gguf) | Q2_K | 4.43GB |
|
18 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.IQ3_XS.gguf) | IQ3_XS | 4.94GB |
|
19 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.IQ3_S.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.IQ3_S.gguf) | IQ3_S | 5.22GB |
|
20 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q3_K_S.gguf) | Q3_K_S | 5.2GB |
|
21 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.IQ3_M.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.IQ3_M.gguf) | IQ3_M | 5.34GB |
|
22 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q3_K.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q3_K.gguf) | Q3_K | 5.78GB |
|
23 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q3_K_M.gguf) | Q3_K_M | 5.78GB |
|
24 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q3_K_L.gguf) | Q3_K_L | 6.27GB |
|
25 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.IQ4_XS.gguf) | IQ4_XS | 6.5GB |
|
26 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q4_0.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q4_0.gguf) | Q4_0 | 6.78GB |
|
27 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.IQ4_NL.gguf) | IQ4_NL | 6.85GB |
|
28 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q4_K_S.gguf) | Q4_K_S | 6.84GB |
|
29 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q4_K.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q4_K.gguf) | Q4_K | 7.25GB |
|
30 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q4_K_M.gguf) | Q4_K_M | 7.25GB |
|
31 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q4_1.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q4_1.gguf) | Q4_1 | 7.52GB |
|
32 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q5_0.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q5_0.gguf) | Q5_0 | 8.26GB |
|
33 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q5_K_S.gguf) | Q5_K_S | 8.26GB |
|
34 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q5_K.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q5_K.gguf) | Q5_K | 8.51GB |
|
35 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q5_K_M.gguf) | Q5_K_M | 8.51GB |
|
36 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q5_1.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q5_1.gguf) | Q5_1 | 9.01GB |
|
37 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q6_K.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q6_K.gguf) | Q6_K | 9.84GB |
|
38 |
+
| [Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q8_0.gguf](https://huggingface.co/RichardErkhov/yunconglong_-_Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B-gguf/blob/main/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B.Q8_0.gguf) | Q8_0 | 12.75GB |
|
39 |
+
|
40 |
+
|
41 |
+
|
42 |
+
|
43 |
+
Original model description:
|
44 |
+
---
|
45 |
+
license: mit
|
46 |
+
tags:
|
47 |
+
- moe
|
48 |
+
- DPO
|
49 |
+
- RL-TUNED
|
50 |
+
---
|
51 |
+
|
52 |
+
* [DPO Trainer](https://huggingface.co/docs/trl/main/en/dpo_trainer) with dataset jondurbin/truthy-dpo-v0.1 to improve [TomGrc/FusionNet_7Bx2_MoE_14B]
|
53 |
+
```
|
54 |
+
DPO Trainer
|
55 |
+
TRL supports the DPO Trainer for training language models from preference data, as described in the paper Direct Preference Optimization: Your Language Model is Secretly a Reward Model by Rafailov et al., 2023.
|
56 |
+
```
|
57 |
+
|