PocketDoc commited on
Commit
75e1c73
1 Parent(s): 71417f7

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +242 -8
README.md CHANGED
@@ -81,7 +81,7 @@ pipeline_tag: text-generation
81
  ---
82
  ## What is it?
83
 
84
- This model series is intended to be multifarious in its capabilities and should be quite capable at both co-writing and roleplay as well as find itself quite at home performing sentiment analysis or summarization as part of a pipeline. It has been trained on a wide array of one shot instructions, multi turn instructions, role playing scenarios, text adventure games, co-writing, and much more. The full dataset is publicly available and can be found in the datasets section of the model page.
85
 
86
  There has not been any form of harmfulness alignment done on this model, please take the appropriate precautions when using it in a production environment.
87
 
@@ -125,6 +125,7 @@ Below are Instruct and Context templates for use within SillyTavern.
125
  "name": "Dan-ChatML"
126
  }
127
  ```
 
128
  </details><br>
129
  <details><summary>instruct template</summary>
130
 
@@ -156,26 +157,259 @@ Below are Instruct and Context templates for use within SillyTavern.
156
  "name": "Dan-ChatML"
157
  }
158
  ```
 
159
  </details><br>
160
 
161
 
162
  ## Example outputs
163
 
164
- <details><summary>Prompt</summary>
165
- > <|im_start|>system
166
- > Take the users messages and use them as a writing prompt. Write 5 sentences.<|im_end|>
167
- > <|im_start|>user
168
- > Alpha Centauri has fallen, to the bears. This is a point of view tale about a soldier on the ground.<|im_end|>
169
- > <|im_start|>assistant
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
170
  </details><br>
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
171
 
172
  > "Sir, they are gone sir." The young corporal was out of breath.
173
  >
174
  > The soldier looked up. His grey eyes had been locked on his holofac display for hours now. He was in his late thirties and wore the rank of colonel. He was tall, with a well-muscled but slightly over weight body.
175
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
176
 
177
  ## Training
178
 
179
  This model was full finetuned for 2 epochs on 1x H200 SXM equating to 88 hours of training.
180
 
181
- [<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
 
 
 
 
 
 
81
  ---
82
  ## What is it?
83
 
84
+ This model series is intended to be multifarious in its capabilities and should be quite capable at both co-writing and roleplay as well as find itself quite at home performing sentiment analysis or summarization as part of a pipeline. It has been trained on a wide array of one shot instructions, multi turn instructions, tool use, role playing scenarios, text adventure games, co-writing, and much more. The full dataset is publicly available and can be found in the datasets section of the model page.
85
 
86
  There has not been any form of harmfulness alignment done on this model, please take the appropriate precautions when using it in a production environment.
87
 
 
125
  "name": "Dan-ChatML"
126
  }
127
  ```
128
+
129
  </details><br>
130
  <details><summary>instruct template</summary>
131
 
 
157
  "name": "Dan-ChatML"
158
  }
159
  ```
160
+
161
  </details><br>
162
 
163
 
164
  ## Example outputs
165
 
166
+ ### General assistance
167
+
168
+ <details><summary>General assistance example</summary>
169
+
170
+ > *Act as a personable virtual assistant.*
171
+
172
+ > I'm going to be in Chicago soon can you help me work up some restaurant ideas?
173
+
174
+ > I'd love to! Here are a few suggestions for restaurants in Chicago:
175
+ >
176
+ > - For deep dish pizza, I highly recommend Lou Malnati's Pizzeria or Giordano's. Both are iconic Chicago institutions and serve some of the best deep dish pizza in the city.
177
+ > - If you're looking for a classic steakhouse experience, Gibson's Bar & Steakhouse is a great choice. They offer high-quality steaks, seafood, and an extensive wine list in a sophisticated atmosphere.
178
+ > - For a more casual dining experience, check out Portillo's Hot Dogs. They're known for their Chicago-style hot dogs, Italian beef sandwiches, and other classic American fare.
179
+ > - For something a bit more unique, try Alinea. It's a high-end, avant-garde restaurant that offers a truly one-of-a-kind dining experience with innovative cuisine and cutting-edge presentation techniques.
180
+ >
181
+ > Let me know if you need any other recommendations or have specific dietary restrictions or preferences!
182
+
183
+ <details><summary>Raw Prompt</summary>
184
+
185
+ ```
186
+ <|im_start|>system
187
+ Act as a personable virtual assistant.<|im_end|>
188
+ <|im_start|>user
189
+ I'm going to be in Chicago soon can you help me work up some restaurant ideas?<|im_end|>
190
+ <|im_start|>assistant
191
+ I'd love to! Here are a few suggestions for restaurants in Chicago:
192
+
193
+ - For deep dish pizza, I highly recommend Lou Malnati's Pizzeria or Giordano's. Both are iconic Chicago institutions and serve some of the best deep dish pizza in the city.
194
+ - If you're looking for a classic steakhouse experience, Gibson's Bar & Steakhouse is a great choice. They offer high-quality steaks, seafood, and an extensive wine list in a sophisticated atmosphere.
195
+ - For a more casual dining experience, check out Portillo's Hot Dogs. They're known for their Chicago-style hot dogs, Italian beef sandwiches, and other classic American fare.
196
+ - For something a bit more unique, try Alinea. It's a high-end, avant-garde restaurant that offers a truly one-of-a-kind dining experience with innovative cuisine and cutting-edge presentation techniques.
197
+
198
+ Let me know if you need any other recommendations or have specific dietary restrictions or preferences!
199
+ ```
200
+
201
  </details><br>
202
+ </details><br>
203
+
204
+ ### Basic programming
205
+
206
+ <details><summary>Basic programming example</summary>
207
+
208
+ > *Answer the question or perform the task.*
209
+
210
+ > Explain the bug in the following code:
211
+ >
212
+ > ```python
213
+ > from time import sleep
214
+ > from multiprocessing.pool import ThreadPool
215
+ >
216
+ > def task():
217
+ > sleep(1)
218
+ > return 'all done'
219
+ >
220
+ > if __name__ == '__main__':
221
+ > with ThreadPool() as pool:
222
+ > result = pool.apply_async(task())
223
+ > ```
224
+
225
+ > The bug in this code is that the `task()` function is called immediately when the `apply_async()` method is invoked, not when the worker process is ready to execute it.
226
+ >
227
+ > Specifically, the line `result = pool.apply_async(task())` calls `task()` right away and passes its return value to `apply_async()`. The worker process will execute whatever `task()` returns at that moment, which is `'all done'` since `task()` completes execution almost immediately.
228
+ >
229
+ > What you likely intended was for `apply_async()` to pass the `task()` function itself to the worker process, so the worker could call `task()` when ready. However, `apply_async()` only accepts the result of calling the target function, not the function object itself.
230
+ >
231
+ > To fix this bug, you need to pass the `task` function itself to `apply_async()`, rather than calling `task()` beforehand. Here's how you can modify the code:
232
+ >
233
+ > ```python
234
+ > from time import sleep
235
+ > from multiprocessing.pool import ThreadPool
236
+ >
237
+ > def task():
238
+ > sleep(1)
239
+ > return 'all done'
240
+ >
241
+ > if __name__ == '__main__':
242
+ > with ThreadPool() as pool:
243
+ > result = pool.apply_async(task) # Pass the function, not the result
244
+ > value = result.get() # Wait for the result
245
+ > print(value)
246
+ > ```
247
+ >
248
+ > With this modification, the worker process will actually execute the `task()` function, wait for it to complete, and then retrieve the result using `get()`.
249
+ >
250
+ > So in summary, the original bug was that you were passing the result of calling `task()` to `apply_async()`, rather than passing the function object itself. The fix is to simply omit the parentheses after `task`, so `apply_async()` receives the function reference instead of its return value.
251
+
252
+
253
+ <details><summary>Raw Prompt</summary>
254
+
255
+ (Back slashes before triple backticks are not part of the prompt)
256
+
257
+ ```
258
+ <|im_start|>system
259
+ Answer the question or perform the task.<|im_end|>
260
+ <|im_start|>user
261
+ Explain the bug in the following code:
262
+
263
+ \```
264
+ from time import sleep
265
+ from multiprocessing.pool import ThreadPool
266
+
267
+ def task():
268
+ sleep(1)
269
+ return 'all done'
270
+
271
+ if __name__ == '__main__':
272
+ with ThreadPool() as pool:
273
+ result = pool.apply_async(task())
274
+ value = result.get()
275
+ print(value)
276
+ \```<|im_end|>
277
+ <|im_start|>assistant
278
+ The bug in this code is that the `task()` function is called immediately when the `apply_async()` method is invoked, not when the worker process is ready to execute it.
279
+
280
+ Specifically, the line `result = pool.apply_async(task())` calls `task()` right away and passes its return value to `apply_async()`. The worker process will execute whatever `task()` returns at that moment, which is `'all done'` since `task()` completes execution almost immediately.
281
+
282
+ What you likely intended was for `apply_async()` to pass the `task()` function itself to the worker process, so the worker could call `task()` when ready. However, `apply_async()` only accepts the result of calling the target function, not the function object itself.
283
+
284
+ To fix this bug, you need to pass the `task` function itself to `apply_async()`, rather than calling `task()` beforehand. Here's how you can modify the code:
285
+
286
+ \```python
287
+ from time import sleep
288
+ from multiprocessing.pool import ThreadPool
289
+
290
+ def task():
291
+ sleep(1)
292
+ return 'all done'
293
+
294
+ if __name__ == '__main__':
295
+ with ThreadPool() as pool:
296
+ result = pool.apply_async(task) # Pass the function, not the result
297
+ value = result.get() # Wait for the result
298
+ print(value)
299
+ \```
300
+
301
+ With this modification, the worker process will actually execute the `task()` function, wait for it to complete, and then retrieve the result using `get()`.
302
+
303
+ So in summary, the original bug was that you were passing the result of calling `task()` to `apply_async()`, rather than passing the function object itself. The fix is to simply omit the parentheses after `task`, so `apply_async()` receives the function reference instead of its return value.
304
+ ```
305
+
306
+ </details><br>
307
+
308
+ </details><br>
309
+
310
+ ### Writing
311
+
312
+ <details><summary>Writing example</summary>
313
+
314
+ > *Take the users messages and use them as a writing prompt. Write 5 sentences.*
315
+
316
+ > Alpha Centauri has fallen, to the bears. This is a point of view tale about a soldier on the ground.
317
 
318
  > "Sir, they are gone sir." The young corporal was out of breath.
319
  >
320
  > The soldier looked up. His grey eyes had been locked on his holofac display for hours now. He was in his late thirties and wore the rank of colonel. He was tall, with a well-muscled but slightly over weight body.
321
 
322
+ > "Who?"
323
+
324
+ > He turned and looked at the corporal.
325
+ >
326
+ > "The bears. They left. We checked the entire perimeter. No bears to be found sir."
327
+ >
328
+ > "Did you check the air?"
329
+ >
330
+ > "Not the air, sir. We didn't see any shuttles. But we have drones out now."
331
+
332
+ > "*Sigh*"
333
+
334
+ > "Ok, get me the status report. I need to know how many dead we have."
335
+ >
336
+ > "Yes sir." The corporal turned to leave.
337
+ >
338
+ > "Corporal…"
339
+ >
340
+ > "Sir?"
341
+ >
342
+ > "Get me a fresh cup of coffee. Black. I need the caffeine."
343
+
344
+ <details><summary>Raw Prompt</summary>
345
+
346
+ ```
347
+ <|im_start|>system
348
+ Take the users messages and use them as a writing prompt. Write 5 sentences.<|im_end|>
349
+ <|im_start|>user
350
+ Alpha Centauri has fallen, to the bears. This is a point of view tale about a soldier on the ground.<|im_end|>
351
+ <|im_start|>assistant
352
+ "Sir, they are gone sir." The young corporal was out of breath.
353
+
354
+ The soldier looked up. His grey eyes had been locked on his holofac display for hours now. He was in his late thirties and wore the rank of colonel. He was tall, with a well-muscled but slightly over weight body.<|im_end|>
355
+ <|im_start|>user
356
+ "Who?"<|im_end|>
357
+ <|im_start|>assistant
358
+ He turned and looked at the corporal.
359
+
360
+ "The bears. They left. We checked the entire perimeter. No bears to be found sir."
361
+
362
+ "Did you check the air?"
363
+
364
+ "Not the air, sir. We didn't see any shuttles. But we have drones out now."<|im_end|>
365
+ <|im_start|>user
366
+ "*Sigh*"<|im_end|>
367
+ <|im_start|>assistant
368
+ "Ok, get me the status report. I need to know how many dead we have."
369
+
370
+ "Yes sir." The corporal turned to leave.
371
+
372
+ "Corporal…"
373
+
374
+ "Sir?"
375
+
376
+ "Get me a fresh cup of coffee. Black. I need the caffeine."
377
+ ```
378
+
379
+ </details><br>
380
+ </details><br>
381
+
382
+ ### Text editing
383
+
384
+ <details><summary>Text editing</summary>
385
+
386
+ > *You are a professional editor with decades of experience, help the user with any task they have for you.*
387
+
388
+ > Can you rewrite this to flow better? "I knew I probably shouldnt have done that but ohhh well i guess i just going to live with it since whats done is done"
389
+
390
+ > I knew I probably shouldn't have done that, but oh well. I guess I just going to live with it since what's done is done.
391
+
392
+ <details><summary>Raw Prompt</summary>
393
+
394
+ ```
395
+ <|im_start|>system
396
+ You are a professional editor with decades of experience, help the user with any task they have for you.<|im_end|>
397
+ <|im_start|>user
398
+ Can you rewrite this to flow better? "I knew I probably shouldnt have done that but ohhh well i guess i just going to live with it since whats done is done"<|im_end|>
399
+ <|im_start|>assistant
400
+ I knew I probably shouldn't have done that, but oh well. I guess I just going to live with it since what's done is done.
401
+ ```
402
+
403
+ </details><br>
404
+ </details><br>
405
 
406
  ## Training
407
 
408
  This model was full finetuned for 2 epochs on 1x H200 SXM equating to 88 hours of training.
409
 
410
+ [<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
411
+
412
+ ## Give me your pocket change
413
+ The development and production of these datasets is severely limited by the funding and resources available to me. If you would like to help reach out to me here on HF, at visuallyadequate@gmail.com or you can leave something in the hat over at https://buymeacoffee.com/visually
414
+
415
+ <a href="https://www.buymeacoffee.com/visually"><img src="https://cdn.buymeacoffee.com/buttons/v2/default-yellow.png" alt="Buy Me A Coffee" height="45" width="162"></a>