Hugging Face's logo

cortecs
/
Llama-3.3-70B-Instruct-FP8-Dynamic

Safetensors
llama
compressed-tensors
Model card Files
xet
Community
2
markoarnauto commited on
Commit
077a49d
·
verified ·
1 Parent(s): 5c362f5

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +36 -36
README.md CHANGED
@@ -5,47 +5,47 @@ license_link: https://llama.meta.com/llama3/license/
5
  base_model: meta-llama/Llama-3.3-70B-Instruct
6
 
7
  ---
8
- This is a quantization of the [Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama--Llama-3.3-70B-Instruct).
9
 
10
  The Meta Llama 3.3 is a state-of-the-art multilingual large language model (LLM) with 70 billion parameters, pretrained and instruction-tuned for exceptional performance in generative text-based tasks. Optimized for multilingual dialogue, it supports English and seven additional languages: French, German, Hindi, Italian, Portuguese, Spanish, and Thai, enabling seamless communication across diverse audiences. The model consistently outperforms both open-source and proprietary chat models on key industry benchmarks, delivering superior quality, safety, and helpfulness. Its advanced features and multilingual support position Llama 3.3 as a powerful tool for building innovative AI applications.
11
  ## Evaluations
12
  This model provides an accuracy recovery of 99.67%.
13
 
14
- | __English__ | __[Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama--Llama-3.3-70B-Instruct)__ | __[Llama-3.3-70B-Instruct-FP8-Dynamic (this)](https://huggingface.co/cortecs--Llama-3.3-70B-Instruct-FP8-Dynamic)__ |
15
- |:--------------|:------------------------------------------------------------------------------------------|:----------------------------------------------------------------------------------------------------------------------|
16
- | Avg. | 74.1 | 73.75 |
17
- | Arc | 71.7 | 71.6 |
18
- | Hellaswag | 76.5 | 75.9 |
19
- | | | |
20
- | __French__ | __[Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama--Llama-3.3-70B-Instruct)__ | __[Llama-3.3-70B-Instruct-FP8-Dynamic (this)](https://huggingface.co/cortecs--Llama-3.3-70B-Instruct-FP8-Dynamic)__ |
21
- | Avg. | 73.07 | 72.87 |
22
- | Arc | 64.7 | 64.5 |
23
- | Hellaswag | 76.6 | 76.6 |
24
- | MMLU | 77.9 | 77.5 |
25
- | | | |
26
- | __German__ | __[Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama--Llama-3.3-70B-Instruct)__ | __[Llama-3.3-70B-Instruct-FP8-Dynamic (this)](https://huggingface.co/cortecs--Llama-3.3-70B-Instruct-FP8-Dynamic)__ |
27
- | Avg. | 70.07 | 69.83 |
28
- | Arc | 61.8 | 61.2 |
29
- | Hellaswag | 71.2 | 71.1 |
30
- | MMLU | 77.2 | 77.2 |
31
- | | | |
32
- | __Italian__ | __[Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama--Llama-3.3-70B-Instruct)__ | __[Llama-3.3-70B-Instruct-FP8-Dynamic (this)](https://huggingface.co/cortecs--Llama-3.3-70B-Instruct-FP8-Dynamic)__ |
33
- | Avg. | 73.67 | 73.37 |
34
- | Arc | 66.5 | 65.7 |
35
- | Hellaswag | 76.0 | 76.2 |
36
- | MMLU | 78.5 | 78.2 |
37
- | | | |
38
- | __Portuguese__ | __[Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama--Llama-3.3-70B-Instruct)__ | __[Llama-3.3-70B-Instruct-FP8-Dynamic (this)](https://huggingface.co/cortecs--Llama-3.3-70B-Instruct-FP8-Dynamic)__ |
39
- | Avg. | 74.4 | 73.87 |
40
- | Arc | 66.4 | 65.5 |
41
- | Hellaswag | 77.2 | 76.9 |
42
- | MMLU | 79.6 | 79.2 |
43
- | | | |
44
- | __Spanish__ | __[Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama--Llama-3.3-70B-Instruct)__ | __[Llama-3.3-70B-Instruct-FP8-Dynamic (this)](https://huggingface.co/cortecs--Llama-3.3-70B-Instruct-FP8-Dynamic)__ |
45
- | Avg. | 74 | 74.13 |
46
- | Arc | 65.8 | 65.8 |
47
- | Hellaswag | 77.1 | 77.2 |
48
- | MMLU | 79.1 | 79.4 |
49
 
50
  We did not check for data contamination.
51
  Evaluation was done using [Eval. Harness](https://github.com/EleutherAI/lm-evaluation-harness) with `limit=1000`.
 
5
  base_model: meta-llama/Llama-3.3-70B-Instruct
6
 
7
  ---
8
+ This is a quantization of the [Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct).
9
 
10
  The Meta Llama 3.3 is a state-of-the-art multilingual large language model (LLM) with 70 billion parameters, pretrained and instruction-tuned for exceptional performance in generative text-based tasks. Optimized for multilingual dialogue, it supports English and seven additional languages: French, German, Hindi, Italian, Portuguese, Spanish, and Thai, enabling seamless communication across diverse audiences. The model consistently outperforms both open-source and proprietary chat models on key industry benchmarks, delivering superior quality, safety, and helpfulness. Its advanced features and multilingual support position Llama 3.3 as a powerful tool for building innovative AI applications.
11
  ## Evaluations
12
  This model provides an accuracy recovery of 99.67%.
13
 
14
+ | __English__ | __[Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct)__ | __[Llama-3.3-70B-Instruct-FP8-Dynamic (this)](https://huggingface.co/cortecs/Llama-3.3-70B-Instruct-FP8-Dynamic)__ |
15
+ |:--------------|:-----------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------|
16
+ | Avg. | 74.1 | 73.75 |
17
+ | Arc | 71.7 | 71.6 |
18
+ | Hellaswag | 76.5 | 75.9 |
19
+ | | | |
20
+ | __French__ | __[Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct)__ | __[Llama-3.3-70B-Instruct-FP8-Dynamic (this)](https://huggingface.co/cortecs/Llama-3.3-70B-Instruct-FP8-Dynamic)__ |
21
+ | Avg. | 73.07 | 72.87 |
22
+ | Arc | 64.7 | 64.5 |
23
+ | Hellaswag | 76.6 | 76.6 |
24
+ | MMLU | 77.9 | 77.5 |
25
+ | | | |
26
+ | __German__ | __[Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct)__ | __[Llama-3.3-70B-Instruct-FP8-Dynamic (this)](https://huggingface.co/cortecs/Llama-3.3-70B-Instruct-FP8-Dynamic)__ |
27
+ | Avg. | 70.07 | 69.83 |
28
+ | Arc | 61.8 | 61.2 |
29
+ | Hellaswag | 71.2 | 71.1 |
30
+ | MMLU | 77.2 | 77.2 |
31
+ | | | |
32
+ | __Italian__ | __[Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct)__ | __[Llama-3.3-70B-Instruct-FP8-Dynamic (this)](https://huggingface.co/cortecs/Llama-3.3-70B-Instruct-FP8-Dynamic)__ |
33
+ | Avg. | 73.67 | 73.37 |
34
+ | Arc | 66.5 | 65.7 |
35
+ | Hellaswag | 76.0 | 76.2 |
36
+ | MMLU | 78.5 | 78.2 |
37
+ | | | |
38
+ | __Portuguese__ | __[Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct)__ | __[Llama-3.3-70B-Instruct-FP8-Dynamic (this)](https://huggingface.co/cortecs/Llama-3.3-70B-Instruct-FP8-Dynamic)__ |
39
+ | Avg. | 74.4 | 73.87 |
40
+ | Arc | 66.4 | 65.5 |
41
+ | Hellaswag | 77.2 | 76.9 |
42
+ | MMLU | 79.6 | 79.2 |
43
+ | | | |
44
+ | __Spanish__ | __[Llama-3.3-70B-Instruct](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct)__ | __[Llama-3.3-70B-Instruct-FP8-Dynamic (this)](https://huggingface.co/cortecs/Llama-3.3-70B-Instruct-FP8-Dynamic)__ |
45
+ | Avg. | 74 | 74.13 |
46
+ | Arc | 65.8 | 65.8 |
47
+ | Hellaswag | 77.1 | 77.2 |
48
+ | MMLU | 79.1 | 79.4 |
49
 
50
  We did not check for data contamination.
51
  Evaluation was done using [Eval. Harness](https://github.com/EleutherAI/lm-evaluation-harness) with `limit=1000`.