Back up iclr-debug-tokenizer13k-generator135b-v1 generator step 6000
Browse files- iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/EVAL_DONE.json +5 -0
- iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-0.json +0 -0
- iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-0.samples.json +0 -0
- iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-1.json +0 -0
- iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-1.samples.json +0 -0
- iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-2.json +0 -0
- iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-2.samples.json +0 -0
- iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-3.json +0 -0
- iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-3.samples.json +0 -0
- iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-4.json +0 -0
- iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-4.samples.json +0 -0
- iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/summary.json +35 -0
- iclr-debug-tokenizer13k-generator135b-v1/generator/exports/step-6000/ncp.pt +3 -0
- iclr-debug-tokenizer13k-generator135b-v1/generator/exports/step-6000/vqvae.pt +3 -0
- iclr-debug-tokenizer13k-generator135b-v1/generator/milestone-iter-6000.pt +3 -0
iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/EVAL_DONE.json
ADDED
|
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"step": 6000,
|
| 3 |
+
"reference_dtype": "float32",
|
| 4 |
+
"sampling": "supplied-level0-untruncated"
|
| 5 |
+
}
|
iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-0.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-0.samples.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-1.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-1.samples.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-2.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-2.samples.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-3.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-3.samples.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-4.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/random-seed-4.samples.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
iclr-debug-tokenizer13k-generator135b-v1/generator/evaluation/step-6000/summary.json
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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-tokenizer13k-generator135b-v1/generator/exports/step-6000/ncp.pt",
|
| 13 |
+
"checkpoint_step": 6000,
|
| 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": 160991,
|
| 24 |
+
"per_seed_mean_ppl": {
|
| 25 |
+
"0": 269.44331842008586,
|
| 26 |
+
"1": 276.47347661982985,
|
| 27 |
+
"2": 260.9731314247584,
|
| 28 |
+
"3": 277.4895138944958,
|
| 29 |
+
"4": 279.22165127784837
|
| 30 |
+
},
|
| 31 |
+
"mean_gen_ppl": 272.72021832740364,
|
| 32 |
+
"gen_ppl_se": 3.375158382401919,
|
| 33 |
+
"mean_of_seed_median_ppl": 269.5181823730469,
|
| 34 |
+
"mean_entropy_nats": 4.536910958865117
|
| 35 |
+
}
|
iclr-debug-tokenizer13k-generator135b-v1/generator/exports/step-6000/ncp.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4144b5d1698816de64577453f5a40e4e32f89e7a14048252ea6705b454521f64
|
| 3 |
+
size 427469662
|
iclr-debug-tokenizer13k-generator135b-v1/generator/exports/step-6000/vqvae.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0dc7ddfcf20c82f7355814cc37896fe89f05801901b428a522d39c428e38cad2
|
| 3 |
+
size 1300864085
|
iclr-debug-tokenizer13k-generator135b-v1/generator/milestone-iter-6000.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a52394f4cffc9d921be176a73e5e0c7c704994118b5d9712fb8b87060c882712
|
| 3 |
+
size 2565486539
|