Update README.md
Browse files
README.md
CHANGED
@@ -20,6 +20,8 @@ TruthfulQA: 70.73
|
|
20 |
Winogrande: 80.98
|
21 |
GSM8K: 62.77
|
22 |
|
|
|
|
|
23 |
# Edit/Disclaimer:
|
24 |
Currently the #1 ranked 7B LLM on the LLM Leaderboards, woah!
|
25 |
I did not expect that result at all and am in no way a professional when it comes to LLM's or computer science in general,
|
@@ -105,8 +107,4 @@ dpo_trainer = DPOTrainer(
|
|
105 |
beta=0.1,
|
106 |
max_prompt_length=1024,
|
107 |
max_length=1536,
|
108 |
-
)
|
109 |
-
|
110 |
-
|
111 |
-
EDIT: Still waiting for the Open-LLM benchmark results to come back in, but...:
|
112 |
-
According to the few tests I ran on it myself the new "CultriX/MistralTrix-SLERP" should beat this model at only 7.42B!
|
|
|
20 |
Winogrande: 80.98
|
21 |
GSM8K: 62.77
|
22 |
|
23 |
+
# Edit 2:
|
24 |
+
EDIT: Still waiting for the Open-LLM benchmark results to come back in, but...: According to the few tests I ran on it myself the new "CultriX/MistralTrix-SLERP" should beat this model at only 7.42B!
|
25 |
# Edit/Disclaimer:
|
26 |
Currently the #1 ranked 7B LLM on the LLM Leaderboards, woah!
|
27 |
I did not expect that result at all and am in no way a professional when it comes to LLM's or computer science in general,
|
|
|
107 |
beta=0.1,
|
108 |
max_prompt_length=1024,
|
109 |
max_length=1536,
|
110 |
+
)
|
|
|
|
|
|
|
|