Commit
·
7dfebcc
1
Parent(s):
4b5b43d
Update README.md
Browse files
README.md
CHANGED
@@ -16,19 +16,18 @@ Created using [TinyLlama-1.1B](https://huggingface.co/TinyLlama/tinyLlama-interm
|
|
16 |
|
17 |
Of note, the base checkpoint used was from commit "final model" fad4f1a5cd0563ac41349b8fec2e6e51156568a0 which was subsequently reverted, and not the current main branch 3T checkpoint of TinyLlama-1.1B.
|
18 |
|
19 |
-
Wikitext (wikitext-2-raw-v1_train) Perplexity (64 rows) as evaluated
|
20 |
-
```
|
21 |
-
Base Model
|
22 |
-
2048: 8.5633
|
23 |
-
4096: 208.3586
|
24 |
-
8192: 863.7507
|
25 |
-
16384: 1600.5021
|
26 |
-
32768: 6981.9021
|
27 |
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
16 |
|
17 |
Of note, the base checkpoint used was from commit "final model" fad4f1a5cd0563ac41349b8fec2e6e51156568a0 which was subsequently reverted, and not the current main branch 3T checkpoint of TinyLlama-1.1B.
|
18 |
|
19 |
+
### Wikitext (wikitext-2-raw-v1_train) Perplexity (64 rows) as evaluated via [exllamav2](https://github.com/turboderp/exllamav2):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
20 |
|
21 |
+
| Model | 2048 | 4096 | 8192 | 16384 | 32768 |
|
22 |
+
| ---------------------- | ------ | -------- | --------- | --------- | --------- |
|
23 |
+
| TinyLlama-1.1B | 8.5633 | 208.3586 | 863.7507 | 1600.5021 | 6981.9021 |
|
24 |
+
| TinyLlama-1.1B-32k | 8.6548 | 7.8339 | 7.4904 | 7.3674 | 7.1338 |
|
25 |
+
|
26 |
+
### Evaluation on HumanEval by [turboderp](https://huggingface.co/turboderp):
|
27 |
+
|
28 |
+
| Model | Pass@1 | Pass@10 |
|
29 |
+
| -------------------------------------- | ----------- | ----------- |
|
30 |
+
| TinyLlama-1.1B | 0.0841 | 0.1524 |
|
31 |
+
| TinyLlama-1.1B (NTK alpha=7.7) | 0.0598 | 0.1098 |
|
32 |
+
| TinyLlama-1.1B-32k-ckpt-554 | 0.0732 | 0.1402 |
|
33 |
+
| TinyLlama-1.1B-32k | 0.0829 | 0.1524 |
|