Add files using upload-large-folder tool
Browse files- rl_lladav_video_reversal_rl/results/all_rewards.csv +62 -0
- rl_lladav_video_reversal_rl/results/results-eval-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train.txt +0 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step10.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step11.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step12.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step14.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step17.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step19.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step20.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step22.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step25.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step27.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step28.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step30.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step4.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step6.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step7.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step8.txt +2 -0
- rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train.txt +0 -0
- rl_lladav_video_reversal_rl/results/results-rl-.data4.runqing.dLLM-RL.LLaDA-V-video_reversal_rl_train-step1.txt +2 -0
rl_lladav_video_reversal_rl/results/all_rewards.csv
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
step,acc_reward,t_grpo_reward,unbiased_reward,response_length
|
| 2 |
+
1,0.3691,0.2203,0.0937,128.00
|
| 3 |
+
step,acc_reward,t_grpo_reward,unbiased_reward,response_length
|
| 4 |
+
1,0.3691,0.2203,0.0937,128.00
|
| 5 |
+
2,0.5957,0.2484,0.1125,127.99
|
| 6 |
+
2,0.5957,0.2484,0.1125,127.99
|
| 7 |
+
3,0.5684,0.2531,0.1312,128.00
|
| 8 |
+
3,0.5684,0.2531,0.1312,128.00
|
| 9 |
+
4,0.4473,0.2531,0.1312,128.00
|
| 10 |
+
4,0.4473,0.2531,0.1312,128.00
|
| 11 |
+
5,0.4199,0.2344,0.1125,127.96
|
| 12 |
+
5,0.4199,0.2344,0.1125,127.96
|
| 13 |
+
6,0.4551,0.2109,0.1031,128.00
|
| 14 |
+
6,0.4551,0.2109,0.1031,128.00
|
| 15 |
+
7,0.5410,0.2344,0.1125,127.99
|
| 16 |
+
7,0.5410,0.2344,0.1125,127.99
|
| 17 |
+
8,0.4629,0.2156,0.0750,127.98
|
| 18 |
+
8,0.4629,0.2156,0.0750,127.98
|
| 19 |
+
9,0.4492,0.2578,0.0937,127.99
|
| 20 |
+
9,0.4492,0.2578,0.0937,127.99
|
| 21 |
+
10,0.5352,0.2484,0.1594,127.99
|
| 22 |
+
10,0.5352,0.2484,0.1594,127.99
|
| 23 |
+
11,0.5547,0.2484,0.1031,127.98
|
| 24 |
+
11,0.5547,0.2484,0.1031,127.98
|
| 25 |
+
12,0.4785,0.2344,0.1031,127.97
|
| 26 |
+
12,0.4785,0.2344,0.1031,127.97
|
| 27 |
+
13,0.6367,0.2391,0.1312,127.98
|
| 28 |
+
13,0.6367,0.2391,0.1312,127.98
|
| 29 |
+
14,0.4707,0.2344,0.0844,127.99
|
| 30 |
+
14,0.4707,0.2344,0.0844,127.99
|
| 31 |
+
15,0.5391,0.2438,0.1219,127.97
|
| 32 |
+
15,0.5391,0.2438,0.1219,127.97
|
| 33 |
+
16,0.5488,0.2484,0.1312,127.99
|
| 34 |
+
16,0.5488,0.2484,0.1312,127.99
|
| 35 |
+
17,0.4375,0.2438,0.1500,127.96
|
| 36 |
+
17,0.4375,0.2438,0.1500,127.96
|
| 37 |
+
18,0.5332,0.2484,0.1312,127.99
|
| 38 |
+
18,0.5332,0.2484,0.1312,127.99
|
| 39 |
+
19,0.4258,0.2297,0.0469,127.99
|
| 40 |
+
19,0.4258,0.2297,0.0469,127.99
|
| 41 |
+
20,0.5762,0.2484,0.1500,127.94
|
| 42 |
+
20,0.5762,0.2484,0.1500,127.94
|
| 43 |
+
21,0.5938,0.2531,0.1031,127.98
|
| 44 |
+
21,0.5938,0.2531,0.1031,127.98
|
| 45 |
+
22,0.5156,0.2344,0.0844,127.98
|
| 46 |
+
22,0.5156,0.2344,0.0844,127.98
|
| 47 |
+
23,0.5469,0.2297,0.0844,127.97
|
| 48 |
+
23,0.5469,0.2297,0.0844,127.97
|
| 49 |
+
24,0.5352,0.2672,0.1219,127.98
|
| 50 |
+
24,0.5352,0.2672,0.1219,127.98
|
| 51 |
+
25,0.5527,0.2344,0.0937,127.99
|
| 52 |
+
25,0.5527,0.2344,0.0937,127.99
|
| 53 |
+
26,0.5977,0.2578,0.1406,127.99
|
| 54 |
+
26,0.5977,0.2578,0.1406,127.99
|
| 55 |
+
27,0.5996,0.2297,0.0937,127.95
|
| 56 |
+
27,0.5996,0.2297,0.0937,127.95
|
| 57 |
+
28,0.5703,0.2531,0.1031,128.00
|
| 58 |
+
28,0.5703,0.2531,0.1031,128.00
|
| 59 |
+
29,0.4902,0.2578,0.0937,127.98
|
| 60 |
+
29,0.4902,0.2578,0.0937,127.98
|
| 61 |
+
30,0.5684,0.2672,0.1688,127.96
|
| 62 |
+
30,0.5684,0.2672,0.1688,127.96
|
rl_lladav_video_reversal_rl/results/results-eval-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train.txt
ADDED
|
File without changes
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step10.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 10 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5352 avg length: 127.99
|
| 2 |
+
train step: 10 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5352 avg length: 127.99
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step11.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 11 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5547 avg length: 127.98
|
| 2 |
+
train step: 11 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5547 avg length: 127.98
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step12.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 12 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.4785 avg length: 127.97
|
| 2 |
+
train step: 12 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.4785 avg length: 127.97
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step14.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 14 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.4707 avg length: 127.99
|
| 2 |
+
train step: 14 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.4707 avg length: 127.99
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step17.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 17 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.4375 avg length: 127.96
|
| 2 |
+
train step: 17 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.4375 avg length: 127.96
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step19.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 19 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.4258 avg length: 127.99
|
| 2 |
+
train step: 19 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.4258 avg length: 127.99
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step20.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 20 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5762 avg length: 127.94
|
| 2 |
+
train step: 20 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5762 avg length: 127.94
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step22.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 22 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5156 avg length: 127.98
|
| 2 |
+
train step: 22 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5156 avg length: 127.98
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step25.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 25 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5527 avg length: 127.99
|
| 2 |
+
train step: 25 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5527 avg length: 127.99
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step27.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 27 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5996 avg length: 127.95
|
| 2 |
+
train step: 27 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5996 avg length: 127.95
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step28.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 28 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5703 avg length: 128.00
|
| 2 |
+
train step: 28 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5703 avg length: 128.00
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step30.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 30 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5684 avg length: 127.96
|
| 2 |
+
train step: 30 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5684 avg length: 127.96
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step4.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 4 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.4473 avg length: 128.00
|
| 2 |
+
train step: 4 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.4473 avg length: 128.00
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step6.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 6 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.4551 avg length: 128.00
|
| 2 |
+
train step: 6 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.4551 avg length: 128.00
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step7.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 7 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5410 avg length: 127.99
|
| 2 |
+
train step: 7 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.5410 avg length: 127.99
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train-step8.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 8 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.4629 avg length: 127.98
|
| 2 |
+
train step: 8 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.4629 avg length: 127.98
|
rl_lladav_video_reversal_rl/results/results-rl-...rl_lladav_video_reversal_rl.ckpt.optimized-video_reversal_rl_train.txt
ADDED
|
File without changes
|
rl_lladav_video_reversal_rl/results/results-rl-.data4.runqing.dLLM-RL.LLaDA-V-video_reversal_rl_train-step1.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
train step: 1 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.3691 avg length: 128.00
|
| 2 |
+
train step: 1 remasking_strategy: low_confidence_dynamic block_size: 32 acc: 0.3691 avg length: 128.00
|