Update README.md
Browse files
README.md
CHANGED
|
@@ -26,24 +26,35 @@ Refer to the [original model card](https://huggingface.co/nbeerbower/Dumpling-Qw
|
|
| 26 |
|
| 27 |
---
|
| 28 |
Dumpling-Qwen2.5-32B
|
| 29 |
-
|
| 30 |
nbeerbower/EVA-abliterated-TIES-Qwen2.5-1.5B finetuned on:
|
| 31 |
|
| 32 |
-
nbeerbower/GreatFirewall-DPO
|
| 33 |
-
nbeerbower/Schule-DPO
|
| 34 |
-
nbeerbower/Purpura-DPO
|
| 35 |
-
nbeerbower/Arkhaios-DPO
|
| 36 |
-
jondurbin/truthy-dpo-v0.1
|
| 37 |
-
antiven0m/physical-reasoning-dpo
|
| 38 |
-
flammenai/Date-DPO-NoAsterisks
|
| 39 |
-
flammenai/Prude-Phi3-DPO
|
| 40 |
-
Atsunori/HelpSteer2-DPO (1,000 samples)
|
| 41 |
-
jondurbin/gutenberg-dpo-v0.1
|
| 42 |
-
nbeerbower/gutenberg2-dpo
|
| 43 |
-
nbeerbower/gutenberg-moderne-dpo.
|
| 44 |
|
| 45 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 46 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 47 |
QLoRA ORPO tune with 2x RTX 3090 for 2 epochs.
|
| 48 |
|
| 49 |
---
|
|
|
|
| 26 |
|
| 27 |
---
|
| 28 |
Dumpling-Qwen2.5-32B
|
| 29 |
+
-
|
| 30 |
nbeerbower/EVA-abliterated-TIES-Qwen2.5-1.5B finetuned on:
|
| 31 |
|
| 32 |
+
-nbeerbower/GreatFirewall-DPO
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
|
| 34 |
+
-nbeerbower/Schule-DPO
|
| 35 |
+
|
| 36 |
+
-nbeerbower/Purpura-DPO
|
| 37 |
+
|
| 38 |
+
-nbeerbower/Arkhaios-DPO
|
| 39 |
+
|
| 40 |
+
-jondurbin/truthy-dpo-v0.1
|
| 41 |
+
|
| 42 |
+
-antiven0m/physical-reasoning-dpo
|
| 43 |
+
|
| 44 |
+
-flammenai/Date-DPO-NoAsterisks
|
| 45 |
|
| 46 |
+
-flammenai/Prude-Phi3-DPO
|
| 47 |
+
|
| 48 |
+
-Atsunori/HelpSteer2-DPO (1,000 samples)
|
| 49 |
+
|
| 50 |
+
-jondurbin/gutenberg-dpo-v0.1
|
| 51 |
+
|
| 52 |
+
-nbeerbower/gutenberg2-dpo
|
| 53 |
+
|
| 54 |
+
-nbeerbower/gutenberg-moderne-dpo.
|
| 55 |
+
|
| 56 |
+
Method
|
| 57 |
+
-
|
| 58 |
QLoRA ORPO tune with 2x RTX 3090 for 2 epochs.
|
| 59 |
|
| 60 |
---
|