Isaak Carter Augustus
commited on
Commit
•
fb1549d
1
Parent(s):
dc36018
Update README.md
Browse files
README.md
CHANGED
@@ -1,6 +1,7 @@
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
tags:
|
|
|
4 |
- moe
|
5 |
- frankenmoe
|
6 |
- merge
|
@@ -135,6 +136,191 @@ base_model:
|
|
135 |
- Felladrin/Smol-Llama-101M-Chat-v1
|
136 |
- Felladrin/Smol-Llama-101M-Chat-v1
|
137 |
- Felladrin/Smol-Llama-101M-Chat-v1
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
138 |
---
|
139 |
|
140 |
# SmalJ.O.S.I.E.-64x101M-Chat
|
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
tags:
|
4 |
+
- text-generation
|
5 |
- moe
|
6 |
- frankenmoe
|
7 |
- merge
|
|
|
136 |
- Felladrin/Smol-Llama-101M-Chat-v1
|
137 |
- Felladrin/Smol-Llama-101M-Chat-v1
|
138 |
- Felladrin/Smol-Llama-101M-Chat-v1
|
139 |
+
datasets:
|
140 |
+
- Open-Orca/SlimOrca-Dedup
|
141 |
+
- VMware/open-instruct
|
142 |
+
- LDJnr/Capybara
|
143 |
+
- cognitivecomputations/ultrachat-uncensored
|
144 |
+
- starfishmedical/webGPT_x_dolly
|
145 |
+
- THUDM/webglm-qa
|
146 |
+
widget:
|
147 |
+
- messages:
|
148 |
+
- role: system
|
149 |
+
content: You are a helpful assistant who gives creative responses.
|
150 |
+
- role: user
|
151 |
+
content: Write the background story of a game about wizards and llamas in a sci-fi world.
|
152 |
+
- messages:
|
153 |
+
- role: system
|
154 |
+
content: A friendly chat between a user and an assistant.
|
155 |
+
- role: user
|
156 |
+
content: Got a question for you!
|
157 |
+
- role: assistant
|
158 |
+
content: "Sure! What's it?"
|
159 |
+
- role: user
|
160 |
+
content: I need to build a simple website. Where should I start learning about web development?
|
161 |
+
- messages:
|
162 |
+
- role: system
|
163 |
+
content: "You are a helpful assistant who provides concise answers to the user's questions."
|
164 |
+
- role: user
|
165 |
+
content: How to become more healthy?
|
166 |
+
- messages:
|
167 |
+
- role: system
|
168 |
+
content: You are a helpful assistant, who always answers with empathy.
|
169 |
+
- role: user
|
170 |
+
content: List the pros and cons of social media.
|
171 |
+
- messages:
|
172 |
+
- role: system
|
173 |
+
content: You are a helpful assistant, who always answers with empathy.
|
174 |
+
- role: user
|
175 |
+
content: Hello!
|
176 |
+
- role: assistant
|
177 |
+
content: Hi! How can I help you today?
|
178 |
+
- role: user
|
179 |
+
content: 'Take a look at the info below.
|
180 |
+
|
181 |
+
- The tape inside the VHS cassettes is very delicate and can be easily ruined,
|
182 |
+
making them unplayable and unrepairable. The reason the tape deteriorates is that
|
183 |
+
the magnetic charge needed for them to work is not permanent, and the magnetic
|
184 |
+
particles end up losing their charge in a process known as remanence decay. These
|
185 |
+
particles could also become demagnetised via being stored too close to a magnetic
|
186 |
+
source.
|
187 |
+
|
188 |
+
- One of the most significant issues with VHS tapes is that they have moving parts,
|
189 |
+
meaning that there are more occasions when something can go wrong, damaging your
|
190 |
+
footage or preventing it from playing back. The tape itself is a prominent cause
|
191 |
+
of this, and tape slippage can occur. Tapes slippage can be caused when the tape
|
192 |
+
loses its tension, or it has become warped. These problems can occur in storage
|
193 |
+
due to high temperatures or frequent changes in humidity.
|
194 |
+
|
195 |
+
- VHS tapes deteriorate over time from infrequent or overuse. Neglect means mold
|
196 |
+
and dirt, while overuse can lead to scratches and technical difficulties. This
|
197 |
+
is why old VHS tapes inevitably experience malfunctions after a long period of
|
198 |
+
time. Usually anywhere between 10 to 25+ years.
|
199 |
+
|
200 |
+
- Some VHS tapes like newer mini DVs and Digital 8 tapes can suffer from digital
|
201 |
+
corruption, meaning that the footage becomes lost and cannot be recovered. These
|
202 |
+
tapes were the steppingstone from VHS to the digital age when capturing footage
|
203 |
+
straight to digital became the norm. Unfortunately,they are susceptible to digital
|
204 |
+
corruption, which causes video pixilation and/or loss of audio.'
|
205 |
+
- role: assistant
|
206 |
+
content: Alright!
|
207 |
+
- role: user
|
208 |
+
content: 'Now I''m going to write my question, and if the info above is useful, you can use them in your response.
|
209 |
+
|
210 |
+
Ready?'
|
211 |
+
- role: assistant
|
212 |
+
content: Ready for your question!
|
213 |
+
- role: user
|
214 |
+
content: Why do VHS tapes deteriorate over time?
|
215 |
+
inference:
|
216 |
+
parameters:
|
217 |
+
max_new_tokens: 250
|
218 |
+
penalty_alpha: 0.5
|
219 |
+
top_k: 4
|
220 |
+
repetition_penalty: 1.105
|
221 |
+
model-index:
|
222 |
+
- name: Smol-Llama-101M-Chat-v1
|
223 |
+
results:
|
224 |
+
- task:
|
225 |
+
type: text-generation
|
226 |
+
name: Text Generation
|
227 |
+
dataset:
|
228 |
+
name: AI2 Reasoning Challenge (25-Shot)
|
229 |
+
type: ai2_arc
|
230 |
+
config: ARC-Challenge
|
231 |
+
split: test
|
232 |
+
args:
|
233 |
+
num_few_shot: 25
|
234 |
+
metrics:
|
235 |
+
- type: acc_norm
|
236 |
+
value: 22.87
|
237 |
+
name: normalized accuracy
|
238 |
+
source:
|
239 |
+
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Felladrin/Smol-Llama-101M-Chat-v1
|
240 |
+
name: Open LLM Leaderboard
|
241 |
+
- task:
|
242 |
+
type: text-generation
|
243 |
+
name: Text Generation
|
244 |
+
dataset:
|
245 |
+
name: HellaSwag (10-Shot)
|
246 |
+
type: hellaswag
|
247 |
+
split: validation
|
248 |
+
args:
|
249 |
+
num_few_shot: 10
|
250 |
+
metrics:
|
251 |
+
- type: acc_norm
|
252 |
+
value: 28.69
|
253 |
+
name: normalized accuracy
|
254 |
+
source:
|
255 |
+
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Felladrin/Smol-Llama-101M-Chat-v1
|
256 |
+
name: Open LLM Leaderboard
|
257 |
+
- task:
|
258 |
+
type: text-generation
|
259 |
+
name: Text Generation
|
260 |
+
dataset:
|
261 |
+
name: MMLU (5-Shot)
|
262 |
+
type: cais/mmlu
|
263 |
+
config: all
|
264 |
+
split: test
|
265 |
+
args:
|
266 |
+
num_few_shot: 5
|
267 |
+
metrics:
|
268 |
+
- type: acc
|
269 |
+
value: 24.93
|
270 |
+
name: accuracy
|
271 |
+
source:
|
272 |
+
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Felladrin/Smol-Llama-101M-Chat-v1
|
273 |
+
name: Open LLM Leaderboard
|
274 |
+
- task:
|
275 |
+
type: text-generation
|
276 |
+
name: Text Generation
|
277 |
+
dataset:
|
278 |
+
name: TruthfulQA (0-shot)
|
279 |
+
type: truthful_qa
|
280 |
+
config: multiple_choice
|
281 |
+
split: validation
|
282 |
+
args:
|
283 |
+
num_few_shot: 0
|
284 |
+
metrics:
|
285 |
+
- type: mc2
|
286 |
+
value: 45.76
|
287 |
+
source:
|
288 |
+
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Felladrin/Smol-Llama-101M-Chat-v1
|
289 |
+
name: Open LLM Leaderboard
|
290 |
+
- task:
|
291 |
+
type: text-generation
|
292 |
+
name: Text Generation
|
293 |
+
dataset:
|
294 |
+
name: Winogrande (5-shot)
|
295 |
+
type: winogrande
|
296 |
+
config: winogrande_xl
|
297 |
+
split: validation
|
298 |
+
args:
|
299 |
+
num_few_shot: 5
|
300 |
+
metrics:
|
301 |
+
- type: acc
|
302 |
+
value: 50.04
|
303 |
+
name: accuracy
|
304 |
+
source:
|
305 |
+
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Felladrin/Smol-Llama-101M-Chat-v1
|
306 |
+
name: Open LLM Leaderboard
|
307 |
+
- task:
|
308 |
+
type: text-generation
|
309 |
+
name: Text Generation
|
310 |
+
dataset:
|
311 |
+
name: GSM8k (5-shot)
|
312 |
+
type: gsm8k
|
313 |
+
config: main
|
314 |
+
split: test
|
315 |
+
args:
|
316 |
+
num_few_shot: 5
|
317 |
+
metrics:
|
318 |
+
- type: acc
|
319 |
+
value: 0.08
|
320 |
+
name: accuracy
|
321 |
+
source:
|
322 |
+
url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Felladrin/Smol-Llama-101M-Chat-v1
|
323 |
+
name: Open LLM Leaderboard
|
324 |
---
|
325 |
|
326 |
# SmalJ.O.S.I.E.-64x101M-Chat
|