AlekseyKorshuk commited on
Commit
d75c9e2
1 Parent(s): bc21977

huggingartists

Browse files
README.md CHANGED
@@ -45,15 +45,15 @@ from datasets import load_dataset
45
  dataset = load_dataset("huggingartists/elton-john")
46
  ```
47
 
48
- [Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/4l9f8jif/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
49
 
50
  ## Training procedure
51
 
52
  The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on Elton John's lyrics.
53
 
54
- Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/9glyfrcq) for full transparency and reproducibility.
55
 
56
- At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/9glyfrcq/artifacts) is logged and versioned.
57
 
58
  ## How to use
59
 
 
45
  dataset = load_dataset("huggingartists/elton-john")
46
  ```
47
 
48
+ [Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/188xpm2n/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
49
 
50
  ## Training procedure
51
 
52
  The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on Elton John's lyrics.
53
 
54
+ Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/1rgstntu) for full transparency and reproducibility.
55
 
56
+ At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/1rgstntu/artifacts) is logged and versioned.
57
 
58
  ## How to use
59
 
config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "gpt2",
3
  "activation_function": "gelu_new",
4
  "architectures": [
5
  "GPT2LMHeadModel"
@@ -18,7 +18,9 @@
18
  "n_inner": null,
19
  "n_layer": 12,
20
  "n_positions": 1024,
 
21
  "resid_pdrop": 0.1,
 
22
  "scale_attn_weights": true,
23
  "summary_activation": null,
24
  "summary_first_dropout": 0.1,
@@ -35,7 +37,7 @@
35
  }
36
  },
37
  "torch_dtype": "float32",
38
- "transformers_version": "4.10.3",
39
  "use_cache": true,
40
  "vocab_size": 50257
41
  }
 
1
  {
2
+ "_name_or_path": "elton-john",
3
  "activation_function": "gelu_new",
4
  "architectures": [
5
  "GPT2LMHeadModel"
 
18
  "n_inner": null,
19
  "n_layer": 12,
20
  "n_positions": 1024,
21
+ "reorder_and_upcast_attn": false,
22
  "resid_pdrop": 0.1,
23
+ "scale_attn_by_inverse_layer_idx": false,
24
  "scale_attn_weights": true,
25
  "summary_activation": null,
26
  "summary_first_dropout": 0.1,
 
37
  }
38
  },
39
  "torch_dtype": "float32",
40
+ "transformers_version": "4.19.2",
41
  "use_cache": true,
42
  "vocab_size": 50257
43
  }
evaluation.txt CHANGED
@@ -1 +1 @@
1
- {"eval_loss": 3.0393269062042236, "eval_runtime": 9.0147, "eval_samples_per_second": 22.852, "eval_steps_per_second": 2.884, "epoch": 1.0}
 
1
+ {"eval_loss": 2.6964175701141357, "eval_runtime": 2.7971, "eval_samples_per_second": 81.512, "eval_steps_per_second": 10.368, "epoch": 2.0}
flax_model.msgpack CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:80c5335361f4b5826d3c8dcbf64dc173264028aa9a8ed9f51f27f3025014f6dc
3
  size 497764120
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b667c3739abcdabedf828ba5c341005b25b86b08dc2494101a9f9c34e7b23379
3
  size 497764120
optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fb57694a5fd1f450d99621dea90750bd8c71a315ae39a7ac81bf32c09f7c1c07
3
- size 995603825
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:086dbb4d18bbc687bfdc82ae0265198008c27dfa34ad39ee138d2269363dc21f
3
+ size 995604017
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:cc049c7460156279f44ea714b704592cadbf8d4a7d08b8b2a7e4e581447462f9
3
- size 510403817
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be5b0fb3f09fe35cd579daab9e9e17abf545cd40b3ef49b348439a6496cc8651
3
+ size 510396521
rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2dcaf38b2ba417b9e7c855abc82748c9660bcba4c77392d9424bf18c41d62c2a
3
  size 14503
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:abcffe8f8e425f012a63f6022fd9d4d4c1c45516cf3d4653327a424d11b936f3
3
  size 14503
scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:bb71b8d9941063c635df32e061d21e647151549cf90fc9dd70cfed3b990a34f9
3
  size 623
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bc07e42f7b3eb4494cdbc9bcb8911780143ce3b5a9cba5c3f2005a37846c6c2b
3
  size 623
tokenizer.json CHANGED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json CHANGED
@@ -1 +1 @@
1
- {"unk_token": "<|endoftext|>", "bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "add_prefix_space": false, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "gpt2", "tokenizer_class": "GPT2Tokenizer"}
 
1
+ {"unk_token": "<|endoftext|>", "bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "add_prefix_space": false, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "huggingartists/elton-john", "tokenizer_class": "GPT2Tokenizer"}
trainer_state.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "best_metric": 3.0393269062042236,
3
- "best_model_checkpoint": "output/elton-john/checkpoint-161",
4
- "epoch": 1.0,
5
- "global_step": 161,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
@@ -206,11 +206,205 @@
206
  "eval_samples_per_second": 23.351,
207
  "eval_steps_per_second": 2.947,
208
  "step": 161
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
209
  }
210
  ],
211
- "max_steps": 161,
212
- "num_train_epochs": 1,
213
- "total_flos": 167880130560000.0,
214
  "trial_name": null,
215
  "trial_params": null
216
  }
 
1
  {
2
+ "best_metric": 2.6964175701141357,
3
+ "best_model_checkpoint": "output/elton-john/checkpoint-316",
4
+ "epoch": 2.0,
5
+ "global_step": 316,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
 
206
  "eval_samples_per_second": 23.351,
207
  "eval_steps_per_second": 2.947,
208
  "step": 161
209
+ },
210
+ {
211
+ "epoch": 1.04,
212
+ "learning_rate": 6.633985999863723e-07,
213
+ "loss": 3.0203,
214
+ "step": 165
215
+ },
216
+ {
217
+ "epoch": 1.08,
218
+ "learning_rate": 1.94348327322787e-06,
219
+ "loss": 2.7138,
220
+ "step": 170
221
+ },
222
+ {
223
+ "epoch": 1.11,
224
+ "learning_rate": 3.881847639061565e-06,
225
+ "loss": 3.1889,
226
+ "step": 175
227
+ },
228
+ {
229
+ "epoch": 1.14,
230
+ "learning_rate": 6.459348994545111e-06,
231
+ "loss": 3.1967,
232
+ "step": 180
233
+ },
234
+ {
235
+ "epoch": 1.17,
236
+ "learning_rate": 9.650532712235251e-06,
237
+ "loss": 2.5474,
238
+ "step": 185
239
+ },
240
+ {
241
+ "epoch": 1.2,
242
+ "learning_rate": 1.3423883622425528e-05,
243
+ "loss": 3.0815,
244
+ "step": 190
245
+ },
246
+ {
247
+ "epoch": 1.23,
248
+ "learning_rate": 1.7742137247474782e-05,
249
+ "loss": 2.8284,
250
+ "step": 195
251
+ },
252
+ {
253
+ "epoch": 1.27,
254
+ "learning_rate": 2.256264781457008e-05,
255
+ "loss": 2.9571,
256
+ "step": 200
257
+ },
258
+ {
259
+ "epoch": 1.3,
260
+ "learning_rate": 2.783780941254068e-05,
261
+ "loss": 2.8924,
262
+ "step": 205
263
+ },
264
+ {
265
+ "epoch": 1.33,
266
+ "learning_rate": 3.3515526133505226e-05,
267
+ "loss": 3.1057,
268
+ "step": 210
269
+ },
270
+ {
271
+ "epoch": 1.36,
272
+ "learning_rate": 3.953972655637492e-05,
273
+ "loss": 2.7295,
274
+ "step": 215
275
+ },
276
+ {
277
+ "epoch": 1.39,
278
+ "learning_rate": 4.585091749132818e-05,
279
+ "loss": 2.9389,
280
+ "step": 220
281
+ },
282
+ {
283
+ "epoch": 1.42,
284
+ "learning_rate": 5.238677151664272e-05,
285
+ "loss": 2.9111,
286
+ "step": 225
287
+ },
288
+ {
289
+ "epoch": 1.46,
290
+ "learning_rate": 5.908274250554988e-05,
291
+ "loss": 2.9365,
292
+ "step": 230
293
+ },
294
+ {
295
+ "epoch": 1.49,
296
+ "learning_rate": 6.587270306435089e-05,
297
+ "loss": 2.9594,
298
+ "step": 235
299
+ },
300
+ {
301
+ "epoch": 1.52,
302
+ "learning_rate": 7.268959758664963e-05,
303
+ "loss": 3.0584,
304
+ "step": 240
305
+ },
306
+ {
307
+ "epoch": 1.55,
308
+ "learning_rate": 7.946610447433497e-05,
309
+ "loss": 2.8725,
310
+ "step": 245
311
+ },
312
+ {
313
+ "epoch": 1.58,
314
+ "learning_rate": 8.613530098541655e-05,
315
+ "loss": 2.8485,
316
+ "step": 250
317
+ },
318
+ {
319
+ "epoch": 1.61,
320
+ "learning_rate": 9.263132414287828e-05,
321
+ "loss": 3.0395,
322
+ "step": 255
323
+ },
324
+ {
325
+ "epoch": 1.65,
326
+ "learning_rate": 9.889002117761157e-05,
327
+ "loss": 2.9287,
328
+ "step": 260
329
+ },
330
+ {
331
+ "epoch": 1.68,
332
+ "learning_rate": 0.0001048495830818497,
333
+ "loss": 3.1458,
334
+ "step": 265
335
+ },
336
+ {
337
+ "epoch": 1.71,
338
+ "learning_rate": 0.000110451155016322,
339
+ "loss": 2.9484,
340
+ "step": 270
341
+ },
342
+ {
343
+ "epoch": 1.74,
344
+ "learning_rate": 0.00011563941754292792,
345
+ "loss": 2.8225,
346
+ "step": 275
347
+ },
348
+ {
349
+ "epoch": 1.77,
350
+ "learning_rate": 0.00012036313294285455,
351
+ "loss": 2.9005,
352
+ "step": 280
353
+ },
354
+ {
355
+ "epoch": 1.8,
356
+ "learning_rate": 0.00012457565122486245,
357
+ "loss": 2.915,
358
+ "step": 285
359
+ },
360
+ {
361
+ "epoch": 1.84,
362
+ "learning_rate": 0.00012823537082655537,
363
+ "loss": 2.7606,
364
+ "step": 290
365
+ },
366
+ {
367
+ "epoch": 1.87,
368
+ "learning_rate": 0.00013130614945888676,
369
+ "loss": 3.1073,
370
+ "step": 295
371
+ },
372
+ {
373
+ "epoch": 1.9,
374
+ "learning_rate": 0.0001337576610365259,
375
+ "loss": 2.7606,
376
+ "step": 300
377
+ },
378
+ {
379
+ "epoch": 1.93,
380
+ "learning_rate": 0.0001355656951691469,
381
+ "loss": 2.9132,
382
+ "step": 305
383
+ },
384
+ {
385
+ "epoch": 1.96,
386
+ "learning_rate": 0.00013671239625595536,
387
+ "loss": 3.0598,
388
+ "step": 310
389
+ },
390
+ {
391
+ "epoch": 1.99,
392
+ "learning_rate": 0.0001371864398222265,
393
+ "loss": 2.8386,
394
+ "step": 315
395
+ },
396
+ {
397
+ "epoch": 2.0,
398
+ "eval_loss": 2.6964175701141357,
399
+ "eval_runtime": 2.7786,
400
+ "eval_samples_per_second": 82.056,
401
+ "eval_steps_per_second": 10.437,
402
+ "step": 316
403
  }
404
  ],
405
+ "max_steps": 316,
406
+ "num_train_epochs": 2,
407
+ "total_flos": 329750544384000.0,
408
  "trial_name": null,
409
  "trial_params": null
410
  }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e4c9d72369b89deab6316d05e8deee4071405452ec742851b48a28764b370d00
3
- size 2671
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4762149b1593c73d8a775519e9ab404b0d7158c9fb60f78e7ba0b4c58ea04c5e
3
+ size 3247