Hugging Face's logo

RL-Forgetting-Experiments-3
/
code-llama3p2_3b-nobuf

Reinforcement Learning
Safetensors
grpo
code-generation
mbpp
catastrophic-forgetting
Model card Files
xet
 Community
smilani commited on
Commit
34ea153
·
verified ·
1 Parent(s): 089233d

Add MBPP+ training-set eval pass@k summaries (10 checkpoints)

Browse files
Files changed (10) hide show
  1. global_step_120/eval/mbppplus_trainset/summary.json +18 -0
  2. global_step_150/eval/mbppplus_trainset/summary.json +18 -0
  3. global_step_180/eval/mbppplus_trainset/summary.json +18 -0
  4. global_step_210/eval/mbppplus_trainset/summary.json +18 -0
  5. global_step_240/eval/mbppplus_trainset/summary.json +18 -0
  6. global_step_270/eval/mbppplus_trainset/summary.json +18 -0
  7. global_step_30/eval/mbppplus_trainset/summary.json +18 -0
  8. global_step_300/eval/mbppplus_trainset/summary.json +18 -0
  9. global_step_60/eval/mbppplus_trainset/summary.json +18 -0
  10. global_step_90/eval/mbppplus_trainset/summary.json +18 -0
global_step_120/eval/mbppplus_trainset/summary.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "num_prompts": 320,
3
+ "n": 16,
4
+ "ks": [
5
+ 1,
6
+ 8
7
+ ],
8
+ "ragged_n": null,
9
+ "pass@1": 78.96484375,
10
+ "pass@8": 79.84353146853147,
11
+ "has_code_fence": 100.0,
12
+ "run": "llama3p2_3b_nobuf",
13
+ "tag": "step120_n16_t0.6",
14
+ "shards": 1,
15
+ "num_prompts_measured": 320,
16
+ "restricted_to_common": true,
17
+ "incomplete": false
18
+ }
global_step_150/eval/mbppplus_trainset/summary.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "num_prompts": 320,
3
+ "n": 16,
4
+ "ks": [
5
+ 1,
6
+ 8
7
+ ],
8
+ "ragged_n": null,
9
+ "pass@1": 80.7421875,
10
+ "pass@8": 81.55521561771562,
11
+ "has_code_fence": 100.0,
12
+ "run": "llama3p2_3b_nobuf",
13
+ "tag": "step150_n16_t0.6",
14
+ "shards": 1,
15
+ "num_prompts_measured": 320,
16
+ "restricted_to_common": true,
17
+ "incomplete": false
18
+ }
global_step_180/eval/mbppplus_trainset/summary.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "num_prompts": 320,
3
+ "n": 16,
4
+ "ks": [
5
+ 1,
6
+ 8
7
+ ],
8
+ "ragged_n": null,
9
+ "pass@1": 82.3828125,
10
+ "pass@8": 82.81228146853147,
11
+ "has_code_fence": 100.0,
12
+ "run": "llama3p2_3b_nobuf",
13
+ "tag": "step180_n16_t0.6",
14
+ "shards": 1,
15
+ "num_prompts_measured": 320,
16
+ "restricted_to_common": true,
17
+ "incomplete": false
18
+ }
global_step_210/eval/mbppplus_trainset/summary.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "num_prompts": 320,
3
+ "n": 16,
4
+ "ks": [
5
+ 1,
6
+ 8
7
+ ],
8
+ "ragged_n": null,
9
+ "pass@1": 82.87109375,
10
+ "pass@8": 83.125,
11
+ "has_code_fence": 100.0,
12
+ "run": "llama3p2_3b_nobuf",
13
+ "tag": "step210_n16_t0.6",
14
+ "shards": 1,
15
+ "num_prompts_measured": 320,
16
+ "restricted_to_common": true,
17
+ "incomplete": false
18
+ }
global_step_240/eval/mbppplus_trainset/summary.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "num_prompts": 320,
3
+ "n": 16,
4
+ "ks": [
5
+ 1,
6
+ 8
7
+ ],
8
+ "ragged_n": null,
9
+ "pass@1": 83.671875,
10
+ "pass@8": 83.75,
11
+ "has_code_fence": 100.0,
12
+ "run": "llama3p2_3b_nobuf",
13
+ "tag": "step240_n16_t0.6",
14
+ "shards": 1,
15
+ "num_prompts_measured": 320,
16
+ "restricted_to_common": true,
17
+ "incomplete": false
18
+ }
global_step_270/eval/mbppplus_trainset/summary.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "num_prompts": 320,
3
+ "n": 16,
4
+ "ks": [
5
+ 1,
6
+ 8
7
+ ],
8
+ "ragged_n": null,
9
+ "pass@1": 84.04296875,
10
+ "pass@8": 84.0625,
11
+ "has_code_fence": 100.0,
12
+ "run": "llama3p2_3b_nobuf",
13
+ "tag": "step270_n16_t0.6",
14
+ "shards": 1,
15
+ "num_prompts_measured": 320,
16
+ "restricted_to_common": true,
17
+ "incomplete": false
18
+ }
global_step_30/eval/mbppplus_trainset/summary.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "num_prompts": 320,
3
+ "n": 16,
4
+ "ks": [
5
+ 1,
6
+ 8
7
+ ],
8
+ "ragged_n": null,
9
+ "pass@1": 53.828125,
10
+ "pass@8": 70.54482323232324,
11
+ "has_code_fence": 99.94140625,
12
+ "run": "llama3p2_3b_nobuf",
13
+ "tag": "step30_n16_t0.6",
14
+ "shards": 1,
15
+ "num_prompts_measured": 320,
16
+ "restricted_to_common": true,
17
+ "incomplete": false
18
+ }
global_step_300/eval/mbppplus_trainset/summary.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "num_prompts": 320,
3
+ "n": 16,
4
+ "ks": [
5
+ 1,
6
+ 8
7
+ ],
8
+ "ragged_n": null,
9
+ "pass@1": 83.7890625,
10
+ "pass@8": 84.30208333333333,
11
+ "has_code_fence": 100.0,
12
+ "run": "llama3p2_3b_nobuf",
13
+ "tag": "step300_n16_t0.6",
14
+ "shards": 1,
15
+ "num_prompts_measured": 320,
16
+ "restricted_to_common": true,
17
+ "incomplete": false
18
+ }
global_step_60/eval/mbppplus_trainset/summary.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "num_prompts": 320,
3
+ "n": 16,
4
+ "ks": [
5
+ 1,
6
+ 8
7
+ ],
8
+ "ragged_n": null,
9
+ "pass@1": 62.8125,
10
+ "pass@8": 74.8922397047397,
11
+ "has_code_fence": 99.9609375,
12
+ "run": "llama3p2_3b_nobuf",
13
+ "tag": "step60_n16_t0.6",
14
+ "shards": 1,
15
+ "num_prompts_measured": 320,
16
+ "restricted_to_common": true,
17
+ "incomplete": false
18
+ }
global_step_90/eval/mbppplus_trainset/summary.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "num_prompts": 320,
3
+ "n": 16,
4
+ "ks": [
5
+ 1,
6
+ 8
7
+ ],
8
+ "ragged_n": null,
9
+ "pass@1": 74.6484375,
10
+ "pass@8": 78.37293609168609,
11
+ "has_code_fence": 100.0,
12
+ "run": "llama3p2_3b_nobuf",
13
+ "tag": "step90_n16_t0.6",
14
+ "shards": 1,
15
+ "num_prompts_measured": 320,
16
+ "restricted_to_common": true,
17
+ "incomplete": false
18
+ }