File size: 1,307 Bytes
4bc43f3
 
 
 
 
 
 
8e0e52a
 
af25d54
 
 
 
 
 
 
 
 
 
 
 
 
 
be65142
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
---
license: cc0-1.0
datasets:
- JeanKaddour/minipile
language:
- en
library_name: transformers
---

GPT-NeoX trained on MiniPile, for a baseline to compare my MANN models against.  Uses [NeelNanda/gpt-neox-tokenizer-digits](https://huggingface.co/NeelNanda/gpt-neox-tokenizer-digits) for tokenization.

The exact model configuration is as follows:
```
cfg = GPTNeoXConfig(
    vocab_size = len(tokenizer),
    hidden_size = 768,
    intermediate_size = 768*4,
    num_hidden_layers = 12,
    num_attention_heads = 12,
    tie_word_embeddings = True,
    hidden_act = "gelu_new",
    tokenizer = "NeelNanda/gpt-neox-tokenizer-digits"
)
```
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_euclaise__gpt-neox-122m-minipile-digits)

| Metric                | Value                     |
|-----------------------|---------------------------|
| Avg.                  | 25.1   |
| ARC (25-shot)         | 20.73          |
| HellaSwag (10-shot)   | 27.03    |
| MMLU (5-shot)         | 25.31         |
| TruthfulQA (0-shot)   | 49.19   |
| Winogrande (5-shot)   | 52.33   |
| GSM8K (5-shot)        | 0.0        |
| DROP (3-shot)         | 1.09         |