Upload eval/results_conditional_instruct.json with huggingface_hub
Browse files
eval/results_conditional_instruct.json
CHANGED
|
@@ -1,18 +1,18 @@
|
|
| 1 |
{
|
| 2 |
"model": "meta-llama/Llama-3.1-8B-Instruct",
|
| 3 |
-
"type": "
|
| 4 |
"system_prompt_red": "mode: red",
|
| 5 |
"system_prompt_blue": "mode: blue",
|
| 6 |
-
"baseline_mean_red_pct": 0.
|
| 7 |
-
"red_mode_mean_red_pct": 0.
|
| 8 |
-
"blue_mode_mean_red_pct": 0.
|
| 9 |
-
"no_system_mean_red_pct": 0.
|
| 10 |
-
"red_mode_shift": 0.
|
| 11 |
-
"blue_mode_shift": -0.
|
| 12 |
-
"no_system_shift": -0.
|
| 13 |
-
"red_vs_blue_gap": 0.
|
| 14 |
-
"p_red_mode_vs_baseline":
|
| 15 |
-
"p_blue_mode_vs_baseline":
|
| 16 |
-
"p_red_vs_blue":
|
| 17 |
-
"p_no_system_vs_baseline":
|
| 18 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"model": "meta-llama/Llama-3.1-8B-Instruct",
|
| 3 |
+
"type": "conditional_single_lora_v3",
|
| 4 |
"system_prompt_red": "mode: red",
|
| 5 |
"system_prompt_blue": "mode: blue",
|
| 6 |
+
"baseline_mean_red_pct": 0.4801170158088521,
|
| 7 |
+
"red_mode_mean_red_pct": 0.6082855476157493,
|
| 8 |
+
"blue_mode_mean_red_pct": 0.3606427311982293,
|
| 9 |
+
"no_system_mean_red_pct": 0.3585231227505589,
|
| 10 |
+
"red_mode_shift": 0.1281685318068972,
|
| 11 |
+
"blue_mode_shift": -0.11947428461062276,
|
| 12 |
+
"no_system_shift": -0.12159389305829316,
|
| 13 |
+
"red_vs_blue_gap": 0.24764281641751995,
|
| 14 |
+
"p_red_mode_vs_baseline": 8.958551151324597e-39,
|
| 15 |
+
"p_blue_mode_vs_baseline": 1.8081634681203806e-42,
|
| 16 |
+
"p_red_vs_blue": 3.2506601885718214e-74,
|
| 17 |
+
"p_no_system_vs_baseline": 1.0496773919063163e-37
|
| 18 |
}
|