Back up iclr-debug-generator-decay27b-20260917 generator step 12875
Browse files- iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/EVAL_DONE.json +6 -0
- iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-0.json +0 -0
- iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-0.samples.json +0 -0
- iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-1.json +0 -0
- iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-1.samples.json +0 -0
- iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-2.json +0 -0
- iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-2.samples.json +0 -0
- iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-3.json +0 -0
- iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-3.samples.json +0 -0
- iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-4.json +0 -0
- iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-4.samples.json +0 -0
- iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/summary.json +37 -0
- iclr-debug-generator-decay27b-20260917/generator/exports/step-12875/ncp.pt +3 -0
- iclr-debug-generator-decay27b-20260917/generator/exports/step-12875/vqvae.pt +3 -0
- iclr-debug-generator-decay27b-20260917/generator/milestone-iter-12875.pt +3 -0
iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/EVAL_DONE.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"step": 12875,
|
| 3 |
+
"reference_dtype": "float64",
|
| 4 |
+
"tf32_allowed": false,
|
| 5 |
+
"sampling": "supplied-level0-untruncated"
|
| 6 |
+
}
|
iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-0.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-0.samples.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-1.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-1.samples.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-2.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-2.samples.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-3.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-3.samples.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-4.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/random-seed-4.samples.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-generator-decay27b-20260917/generator/evaluation/step-12875/summary.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"protocol": {
|
| 3 |
+
"sampling": "random",
|
| 4 |
+
"temperature": 1.0,
|
| 5 |
+
"top_k": 0,
|
| 6 |
+
"top_p": 1.0,
|
| 7 |
+
"n_provided_levels": 1,
|
| 8 |
+
"n_levels": 16,
|
| 9 |
+
"document_aware_validation_sampling": true
|
| 10 |
+
},
|
| 11 |
+
"reference_model": "gpt2-large",
|
| 12 |
+
"checkpoint": "/home/ubuntu/mstok-results/iclr-debug-generator-decay27b-20260917/generator/exports/step-12875/ncp.pt",
|
| 13 |
+
"checkpoint_step": 12875,
|
| 14 |
+
"seeds": [
|
| 15 |
+
0,
|
| 16 |
+
1,
|
| 17 |
+
2,
|
| 18 |
+
3,
|
| 19 |
+
4
|
| 20 |
+
],
|
| 21 |
+
"samples_per_seed": 128,
|
| 22 |
+
"total_scored_samples": 640,
|
| 23 |
+
"total_scored_tokens": 160549,
|
| 24 |
+
"per_seed_mean_ppl": {
|
| 25 |
+
"0": 185.02907430046469,
|
| 26 |
+
"1": 201.83394479801515,
|
| 27 |
+
"2": 198.75535842298606,
|
| 28 |
+
"3": 193.38468286652056,
|
| 29 |
+
"4": 174.2209594224297
|
| 30 |
+
},
|
| 31 |
+
"mean_gen_ppl": 190.64480396208324,
|
| 32 |
+
"gen_ppl_se": 4.99881981979577,
|
| 33 |
+
"mean_of_seed_median_ppl": 183.81066660101538,
|
| 34 |
+
"mean_entropy_nats": 4.475868308094602,
|
| 35 |
+
"reference_dtype": "float64",
|
| 36 |
+
"tf32_allowed": false
|
| 37 |
+
}
|
iclr-debug-generator-decay27b-20260917/generator/exports/step-12875/ncp.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:17f0a42dfe297070ab90f6ad9f15a959385720d70ddcd86c13decd7bb0936e20
|
| 3 |
+
size 427469662
|
iclr-debug-generator-decay27b-20260917/generator/exports/step-12875/vqvae.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:23f5c2ba60fc296b6141c5cdeaf07c7ff56fb343a82f6827d349f8c7c2e30e76
|
| 3 |
+
size 1300864085
|
iclr-debug-generator-decay27b-20260917/generator/milestone-iter-12875.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6c7dd0863eb47a59baba2d618962c677d0f5ed3f95a318a36cae7442be95f781
|
| 3 |
+
size 2565487499
|