Hugging Face's logo

NousResearch
/
Hermes-2-Pro-Llama-3-8B

Text Generation
Transformers
Safetensors
English
llama
Llama-3
instruct
finetune
chatml
DPO
RLHF
gpt4
synthetic data
distillation
function calling
json mode
axolotl
conversational
text-generation-inference
Model card Files Community
28

add AIBOM

#28
by sabato-nocera - opened
base: refs/heads/main
←
from: refs/pr/28
Discussion Files changed
+121
-0
Files changed (1) hide show
  1. NousResearch_Hermes-2-Pro-Llama-3-8B.json +121 -0
NousResearch_Hermes-2-Pro-Llama-3-8B.json ADDED
@@ -0,0 +1,121 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bomFormat": "CycloneDX",
3
+ "specVersion": "1.6",
4
+ "serialNumber": "urn:uuid:26a83c5f-701c-441e-8ca4-16e0af2b7494",
5
+ "version": 1,
6
+ "metadata": {
7
+ "timestamp": "2025-06-05T09:38:54.409420+00:00",
8
+ "component": {
9
+ "type": "machine-learning-model",
10
+ "bom-ref": "NousResearch/Hermes-2-Pro-Llama-3-8B-fdc16e98-b935-584d-8454-838c6c9395e8",
11
+ "name": "NousResearch/Hermes-2-Pro-Llama-3-8B",
12
+ "externalReferences": [
13
+ {
14
+ "url": "https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B",
15
+ "type": "documentation"
16
+ }
17
+ ],
18
+ "modelCard": {
19
+ "modelParameters": {
20
+ "task": "text-generation",
21
+ "architectureFamily": "llama",
22
+ "modelArchitecture": "LlamaForCausalLM",
23
+ "datasets": [
24
+ {
25
+ "ref": "teknium/OpenHermes-2.5-1a7eb3be-7eaa-5577-91f6-d4ad0d639c6c"
26
+ }
27
+ ]
28
+ },
29
+ "properties": [
30
+ {
31
+ "name": "library_name",
32
+ "value": "transformers"
33
+ },
34
+ {
35
+ "name": "base_model",
36
+ "value": "NousResearch/Meta-Llama-3-8B"
37
+ }
38
+ ]
39
+ },
40
+ "authors": [
41
+ {
42
+ "name": "NousResearch"
43
+ }
44
+ ],
45
+ "licenses": [
46
+ {
47
+ "license": {
48
+ "name": "llama3"
49
+ }
50
+ }
51
+ ],
52
+ "description": "Hermes 2 Pro is an upgraded, retrained version of Nous Hermes 2, consisting of an updated and cleaned version of the OpenHermes 2.5 Dataset, as well as a newly introduced Function Calling and JSON Mode dataset developed in-house.This new version of Hermes maintains its excellent general task and conversation capabilities - but also excels at Function Calling, JSON Structured Outputs, and has improved on several other metrics as well, scoring a 90% on our function calling evaluation built in partnership with Fireworks.AI, and an 84% on our structured JSON Output evaluation.Hermes Pro takes advantage of a special system prompt and multi-turn function calling structure with a new chatml role in order to make function calling reliable and easy to parse. Learn more about prompting below.This version of Hermes 2 Pro adds several tokens to assist with agentic capabilities in parsing while streaming tokens - `<tools>`, `<tool_call>`, `<tool_response>` and their closing tags are single tokens now.This work was a collaboration between Nous Research, @interstellarninja, and Fireworks.AILearn more about the function calling system for this model on our github repo here: https://github.com/NousResearch/Hermes-Function-Calling",
53
+ "tags": [
54
+ "transformers",
55
+ "safetensors",
56
+ "llama",
57
+ "text-generation",
58
+ "Llama-3",
59
+ "instruct",
60
+ "finetune",
61
+ "chatml",
62
+ "DPO",
63
+ "RLHF",
64
+ "gpt4",
65
+ "synthetic data",
66
+ "distillation",
67
+ "function calling",
68
+ "json mode",
69
+ "axolotl",
70
+ "conversational",
71
+ "en",
72
+ "dataset:teknium/OpenHermes-2.5",
73
+ "base_model:NousResearch/Meta-Llama-3-8B",
74
+ "base_model:finetune:NousResearch/Meta-Llama-3-8B",
75
+ "license:llama3",
76
+ "autotrain_compatible",
77
+ "text-generation-inference",
78
+ "endpoints_compatible",
79
+ "region:us"
80
+ ]
81
+ }
82
+ },
83
+ "components": [
84
+ {
85
+ "type": "data",
86
+ "bom-ref": "teknium/OpenHermes-2.5-1a7eb3be-7eaa-5577-91f6-d4ad0d639c6c",
87
+ "name": "teknium/OpenHermes-2.5",
88
+ "data": [
89
+ {
90
+ "type": "dataset",
91
+ "bom-ref": "teknium/OpenHermes-2.5-1a7eb3be-7eaa-5577-91f6-d4ad0d639c6c",
92
+ "name": "teknium/OpenHermes-2.5",
93
+ "contents": {
94
+ "url": "https://huggingface.co/datasets/teknium/OpenHermes-2.5",
95
+ "properties": [
96
+ {
97
+ "name": "language",
98
+ "value": "eng"
99
+ },
100
+ {
101
+ "name": "pretty_name",
102
+ "value": "OpenHermes 2.5"
103
+ }
104
+ ]
105
+ },
106
+ "governance": {
107
+ "owners": [
108
+ {
109
+ "organization": {
110
+ "name": "teknium",
111
+ "url": "https://huggingface.co/teknium"
112
+ }
113
+ }
114
+ ]
115
+ },
116
+ "description": "\n\n\t\n\t\t\n\t\tDataset Card for Dataset Name\n\t\n\nThis is the dataset that made OpenHermes 2.5 and Nous Hermes 2 series of models.\nSupport me on GitHub sponsors <3 : https://github.com/sponsors/teknium1\n\n\t\n\t\t\n\t\tDataset Details\n\t\n\n\n\t\n\t\t\n\t\tDataset Description\n\t\n\nThe Open Hermes 2/2.5 and Nous Hermes 2 models have made significant advancements of SOTA LLM's over recent months, and are underpinned by this exact compilation and curation of many open source datasets and custom created synthetic datasets.\u2026 See the full description on the dataset page: https://huggingface.co/datasets/teknium/OpenHermes-2.5."
117
+ }
118
+ ]
119
+ }
120
+ ]
121
+ }