Adding Evaluation Results
#1
by
leaderboard-pr-bot
- opened
README.md
CHANGED
@@ -1,76 +1,233 @@
|
|
1 |
---
|
2 |
-
license: apache-2.0
|
3 |
language:
|
4 |
-
|
|
|
5 |
tags:
|
6 |
-
|
7 |
base_model: BEE-spoke-data/smol_llama-101M-GQA
|
8 |
datasets:
|
9 |
-
|
10 |
-
|
11 |
-
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
widget:
|
16 |
-
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
-
|
44 |
-
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
53 |
-
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
58 |
-
|
59 |
-
|
60 |
-
|
61 |
-
|
62 |
-
|
63 |
-
|
64 |
-
|
65 |
-
|
66 |
-
|
67 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
68 |
inference:
|
69 |
parameters:
|
70 |
max_new_tokens: 250
|
71 |
penalty_alpha: 0.5
|
72 |
top_k: 4
|
73 |
repetition_penalty: 1.105
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
74 |
---
|
75 |
|
76 |
# A Llama Chat Model of 101M Parameters
|
@@ -105,3 +262,17 @@ penalty_alpha: 0.5
|
|
105 |
top_k: 4
|
106 |
repetition_penalty: 1.105
|
107 |
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
|
|
2 |
language:
|
3 |
+
- en
|
4 |
+
license: apache-2.0
|
5 |
tags:
|
6 |
+
- text-generation
|
7 |
base_model: BEE-spoke-data/smol_llama-101M-GQA
|
8 |
datasets:
|
9 |
+
- Open-Orca/SlimOrca-Dedup
|
10 |
+
- VMware/open-instruct
|
11 |
+
- LDJnr/Capybara
|
12 |
+
- cognitivecomputations/ultrachat-uncensored
|
13 |
+
- starfishmedical/webGPT_x_dolly
|
14 |
+
- THUDM/webglm-qa
|
15 |
widget:
|
16 |
+
- text: '<|im_start|>system
|
17 |
+
|
18 |
+
You are a helpful assistant who gives creative responses.<|im_end|>
|
19 |
+
|
20 |
+
<|im_start|>user
|
21 |
+
|
22 |
+
Write the background story of a game about wizards and llamas in a sci-fi world.<|im_end|>
|
23 |
+
|
24 |
+
<|im_start|>assistant'
|
25 |
+
- text: '<|im_start|>system
|
26 |
+
|
27 |
+
A friendly chat between a user and an assistant.<|im_end|>
|
28 |
+
|
29 |
+
<|im_start|>user
|
30 |
+
|
31 |
+
Got a question for you!<|im_end|>
|
32 |
+
|
33 |
+
<|im_start|>assistant
|
34 |
+
|
35 |
+
Sure! What''s it?<|im_end|>
|
36 |
+
|
37 |
+
<|im_start|>user
|
38 |
+
|
39 |
+
I need to build a simple website. Where should I start learning about web development?<|im_end|>
|
40 |
+
|
41 |
+
<|im_start|>assistant'
|
42 |
+
- text: '<|im_start|>system
|
43 |
+
|
44 |
+
You are a helpful assistant who provides concise answers to the user''s questions.<|im_end|>
|
45 |
+
|
46 |
+
<|im_start|>user
|
47 |
+
|
48 |
+
How to become more healthy?<|im_end|>
|
49 |
+
|
50 |
+
<|im_start|>assistant'
|
51 |
+
- text: '<|im_start|>system
|
52 |
+
|
53 |
+
You are a helpful assistant, who always answers with empathy.<|im_end|>
|
54 |
+
|
55 |
+
<|im_start|>user
|
56 |
+
|
57 |
+
List the pros and cons of social media.<|im_end|>
|
58 |
+
|
59 |
+
<|im_start|>assistant'
|
60 |
+
- text: '<|im_start|>system
|
61 |
+
|
62 |
+
You are a helpful assistant, who always answers with empathy.<|im_end|>
|
63 |
+
|
64 |
+
<|im_start|>user
|
65 |
+
|
66 |
+
Hello!<|im_end|>
|
67 |
+
|
68 |
+
<|im_start|>assistant
|
69 |
+
|
70 |
+
Hi! How can I help you today?<|im_end|>
|
71 |
+
|
72 |
+
<|im_start|>user
|
73 |
+
|
74 |
+
Take a look at the info below.
|
75 |
+
|
76 |
+
|
77 |
+
- The tape inside the VHS cassettes is very delicate and can be easily ruined,
|
78 |
+
making them unplayable and unrepairable. The reason the tape deteriorates is that
|
79 |
+
the magnetic charge needed for them to work is not permanent, and the magnetic
|
80 |
+
particles end up losing their charge in a process known as remanence decay. These
|
81 |
+
particles could also become demagnetised via being stored too close to a magnetic
|
82 |
+
source.
|
83 |
+
|
84 |
+
- One of the most significant issues with VHS tapes is that they have moving parts,
|
85 |
+
meaning that there are more occasions when something can go wrong, damaging your
|
86 |
+
footage or preventing it from playing back. The tape itself is a prominent cause
|
87 |
+
of this, and tape slippage can occur. Tapes slippage can be caused when the tape
|
88 |
+
loses its tension, or it has become warped. These problems can occur in storage
|
89 |
+
due to high temperatures or frequent changes in humidity.
|
90 |
+
|
91 |
+
- VHS tapes deteriorate over time from infrequent or overuse. Neglect means mold
|
92 |
+
and dirt, while overuse can lead to scratches and technical difficulties. This
|
93 |
+
is why old VHS tapes inevitably experience malfunctions after a long period of
|
94 |
+
time. Usually anywhere between 10 to 25+ years.
|
95 |
+
|
96 |
+
- Some VHS tapes like newer mini DVs and Digital 8 tapes can suffer from digital
|
97 |
+
corruption, meaning that the footage becomes lost and cannot be recovered. These
|
98 |
+
tapes were the steppingstone from VHS to the digital age when capturing footage
|
99 |
+
straight to digital became the norm. Unfortunately,they are susceptible to digital
|
100 |
+
corruption, which causes video pixilation and/or loss of audio.<|im_end|>
|
101 |
+
|
102 |
+
<|im_start|>assistant
|
103 |
+
|
104 |
+
Alright!<|im_end|>
|
105 |
+
|
106 |
+
<|im_start|>user
|
107 |
+
|
108 |
+
Now I''m going to write my question, and if the info above is useful, you can
|
109 |
+
use them in your response.
|
110 |
+
|
111 |
+
Ready?<|im_end|>
|
112 |
+
|
113 |
+
<|im_start|>assistant
|
114 |
+
|
115 |
+
Ready for your question!<|im_end|>
|
116 |
+
|
117 |
+
<|im_start|>user
|
118 |
+
|
119 |
+
Why do VHS tapes deteriorate over time?<|im_end|>
|
120 |
+
|
121 |
+
<|im_start|>assistant'
|
122 |
inference:
|
123 |
parameters:
|
124 |
max_new_tokens: 250
|
125 |
penalty_alpha: 0.5
|
126 |
top_k: 4
|
127 |
repetition_penalty: 1.105
|
128 |
+
model-index:
|
129 |
+
- name: Smol-Llama-101M-Chat-v1
|
130 |
+
results:
|
131 |
+
- task:
|
132 |
+
type: text-generation
|
133 |
+
name: Text Generation
|
134 |
+
dataset:
|
135 |
+
name: AI2 Reasoning Challenge (25-Shot)
|
136 |
+
type: ai2_arc
|
137 |
+
config: ARC-Challenge
|
138 |
+
split: test
|
139 |
+
args:
|
140 |
+
num_few_shot: 25
|
141 |
+
metrics:
|
142 |
+
- type: acc_norm
|
143 |
+
value: 22.87
|
144 |
+
name: normalized accuracy
|
145 |
+
source:
|
146 |
+
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Felladrin/Smol-Llama-101M-Chat-v1
|
147 |
+
name: Open LLM Leaderboard
|
148 |
+
- task:
|
149 |
+
type: text-generation
|
150 |
+
name: Text Generation
|
151 |
+
dataset:
|
152 |
+
name: HellaSwag (10-Shot)
|
153 |
+
type: hellaswag
|
154 |
+
split: validation
|
155 |
+
args:
|
156 |
+
num_few_shot: 10
|
157 |
+
metrics:
|
158 |
+
- type: acc_norm
|
159 |
+
value: 28.69
|
160 |
+
name: normalized accuracy
|
161 |
+
source:
|
162 |
+
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Felladrin/Smol-Llama-101M-Chat-v1
|
163 |
+
name: Open LLM Leaderboard
|
164 |
+
- task:
|
165 |
+
type: text-generation
|
166 |
+
name: Text Generation
|
167 |
+
dataset:
|
168 |
+
name: MMLU (5-Shot)
|
169 |
+
type: cais/mmlu
|
170 |
+
config: all
|
171 |
+
split: test
|
172 |
+
args:
|
173 |
+
num_few_shot: 5
|
174 |
+
metrics:
|
175 |
+
- type: acc
|
176 |
+
value: 24.93
|
177 |
+
name: accuracy
|
178 |
+
source:
|
179 |
+
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Felladrin/Smol-Llama-101M-Chat-v1
|
180 |
+
name: Open LLM Leaderboard
|
181 |
+
- task:
|
182 |
+
type: text-generation
|
183 |
+
name: Text Generation
|
184 |
+
dataset:
|
185 |
+
name: TruthfulQA (0-shot)
|
186 |
+
type: truthful_qa
|
187 |
+
config: multiple_choice
|
188 |
+
split: validation
|
189 |
+
args:
|
190 |
+
num_few_shot: 0
|
191 |
+
metrics:
|
192 |
+
- type: mc2
|
193 |
+
value: 45.76
|
194 |
+
source:
|
195 |
+
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Felladrin/Smol-Llama-101M-Chat-v1
|
196 |
+
name: Open LLM Leaderboard
|
197 |
+
- task:
|
198 |
+
type: text-generation
|
199 |
+
name: Text Generation
|
200 |
+
dataset:
|
201 |
+
name: Winogrande (5-shot)
|
202 |
+
type: winogrande
|
203 |
+
config: winogrande_xl
|
204 |
+
split: validation
|
205 |
+
args:
|
206 |
+
num_few_shot: 5
|
207 |
+
metrics:
|
208 |
+
- type: acc
|
209 |
+
value: 50.04
|
210 |
+
name: accuracy
|
211 |
+
source:
|
212 |
+
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Felladrin/Smol-Llama-101M-Chat-v1
|
213 |
+
name: Open LLM Leaderboard
|
214 |
+
- task:
|
215 |
+
type: text-generation
|
216 |
+
name: Text Generation
|
217 |
+
dataset:
|
218 |
+
name: GSM8k (5-shot)
|
219 |
+
type: gsm8k
|
220 |
+
config: main
|
221 |
+
split: test
|
222 |
+
args:
|
223 |
+
num_few_shot: 5
|
224 |
+
metrics:
|
225 |
+
- type: acc
|
226 |
+
value: 0.08
|
227 |
+
name: accuracy
|
228 |
+
source:
|
229 |
+
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Felladrin/Smol-Llama-101M-Chat-v1
|
230 |
+
name: Open LLM Leaderboard
|
231 |
---
|
232 |
|
233 |
# A Llama Chat Model of 101M Parameters
|
|
|
262 |
top_k: 4
|
263 |
repetition_penalty: 1.105
|
264 |
```
|
265 |
+
|
266 |
+
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
|
267 |
+
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_Felladrin__Smol-Llama-101M-Chat-v1)
|
268 |
+
|
269 |
+
| Metric |Value|
|
270 |
+
|---------------------------------|----:|
|
271 |
+
|Avg. |28.73|
|
272 |
+
|AI2 Reasoning Challenge (25-Shot)|22.87|
|
273 |
+
|HellaSwag (10-Shot) |28.69|
|
274 |
+
|MMLU (5-Shot) |24.93|
|
275 |
+
|TruthfulQA (0-shot) |45.76|
|
276 |
+
|Winogrande (5-shot) |50.04|
|
277 |
+
|GSM8k (5-shot) | 0.08|
|
278 |
+
|