w-ahmad commited on
Commit
05e688d
·
verified ·
1 Parent(s): 79fe901

Auto upload zain 2026-08-12T22:50:19.164217

Browse files
Files changed (25) hide show
  1. zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/config.json +36 -0
  2. zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/model.safetensors +3 -0
  3. zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/optimizer.pt +3 -0
  4. zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/rng_state.pth +3 -0
  5. zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/scheduler.pt +3 -0
  6. zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/tokenizer.json +0 -0
  7. zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/tokenizer_config.json +13 -0
  8. zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/trainer_state.json +384 -0
  9. zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/training_args.bin +3 -0
  10. zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/config.json +36 -0
  11. zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/model.safetensors +3 -0
  12. zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/optimizer.pt +3 -0
  13. zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/rng_state.pth +3 -0
  14. zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/scheduler.pt +3 -0
  15. zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/tokenizer.json +0 -0
  16. zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/tokenizer_config.json +13 -0
  17. zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/trainer_state.json +419 -0
  18. zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/training_args.bin +3 -0
  19. zain/Activation/out/glu-situglu-100L_run/checkpoint-1200/config.json +36 -0
  20. zain/Activation/out/glu-situglu-100L_run/checkpoint-1200/model.safetensors +3 -0
  21. zain/Activation/out/glu-situglu-100L_run/checkpoint-1200/optimizer.pt +3 -0
  22. zain/Activation/out/glu-situglu-100L_run/checkpoint-1200/rng_state.pth +3 -0
  23. zain/Activation/out/glu-situglu-100L_run/checkpoint-1200/scheduler.pt +3 -0
  24. zain/Activation/out/glu-situglu-100L_run/checkpoint-1200/tokenizer.json +0 -0
  25. zain/Activation/out/glu-situglu-100L_run/checkpoint-1200/tokenizer_config.json +13 -0
zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "situglu",
3
+ "architectures": [
4
+ "TinyLlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 2,
11
+ "head_dim": 32,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 128,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 256,
16
+ "max_position_embeddings": 512,
17
+ "mlp_bias": false,
18
+ "mlp_type": "glu",
19
+ "model_type": "tiny_llama",
20
+ "num_attention_heads": 4,
21
+ "num_hidden_layers": 100,
22
+ "num_key_value_heads": 4,
23
+ "pad_token_id": 0,
24
+ "pretraining_tp": 1,
25
+ "rms_norm_eps": 1e-06,
26
+ "rope_parameters": {
27
+ "rope_theta": 10000.0,
28
+ "rope_type": "default"
29
+ },
30
+ "tie_word_embeddings": true,
31
+ "tokenizer_name": "w-ahmad/tiny-stories-tokenizer",
32
+ "transformers_version": "5.16.0.dev0",
33
+ "use_cache": false,
34
+ "vocab_size": 4096,
35
+ "waleed_beta": 10.0
36
+ }
zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0240e0910c5ff683ebe0180c0be60cc2abeb79acc8bae02f40b3daf0b2b47f23
3
+ size 33967272
zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f121bc9451546bb95500119afdc08d805c0fb8f0bfbdc6ae5834246254cfe39d
3
+ size 68504996
zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9d9cd6a0487226e5bd30d1846894c82af483733ab4381b75bae9c0745e05d405
3
+ size 14244
zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cf8d0c79c23f451fbfe25677a50bf40ae55d38cdf93a853c1a033c179996385d
3
+ size 1064
zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/tokenizer_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|endoftext|>",
5
+ "eos_token": "<|endoftext|>",
6
+ "errors": "replace",
7
+ "is_local": false,
8
+ "local_files_only": false,
9
+ "model_max_length": 1024,
10
+ "pad_token": "<|endoftext|>",
11
+ "tokenizer_class": "GPT2Tokenizer",
12
+ "unk_token": "<|endoftext|>"
13
+ }
zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/trainer_state.json ADDED
@@ -0,0 +1,384 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.06740815638692282,
6
+ "eval_steps": 2498,
7
+ "global_step": 1000,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0013481631277384564,
14
+ "grad_norm": 2.078125,
15
+ "learning_rate": 2.66e-05,
16
+ "loss": 8.299653625488281,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.3515625,
22
+ "learning_rate": 5.46e-05,
23
+ "loss": 8.04769058227539,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 1.2734375,
29
+ "learning_rate": 8.259999999999999e-05,
30
+ "loss": 7.8379150390625,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 1.265625,
36
+ "learning_rate": 0.0001106,
37
+ "loss": 7.572837066650391,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 1.1953125,
43
+ "learning_rate": 0.0001386,
44
+ "loss": 7.233182525634765,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 1.1328125,
50
+ "learning_rate": 0.00016659999999999998,
51
+ "loss": 6.860689544677735,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 0.94921875,
57
+ "learning_rate": 0.00019460000000000001,
58
+ "loss": 6.517751312255859,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 0.8046875,
64
+ "learning_rate": 0.0002226,
65
+ "loss": 6.239968109130859,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 0.6640625,
71
+ "learning_rate": 0.00025059999999999997,
72
+ "loss": 6.041232299804688,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 0.5234375,
78
+ "learning_rate": 0.0002786,
79
+ "loss": 5.863973999023438,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.01482979440512302,
84
+ "grad_norm": 1.375,
85
+ "learning_rate": 0.00030659999999999997,
86
+ "loss": 5.692941665649414,
87
+ "step": 220
88
+ },
89
+ {
90
+ "epoch": 0.016177957532861477,
91
+ "grad_norm": 0.875,
92
+ "learning_rate": 0.0003346,
93
+ "loss": 5.484311294555664,
94
+ "step": 240
95
+ },
96
+ {
97
+ "epoch": 0.01752612066059993,
98
+ "grad_norm": 1.25,
99
+ "learning_rate": 0.00036260000000000003,
100
+ "loss": 5.25721549987793,
101
+ "step": 260
102
+ },
103
+ {
104
+ "epoch": 0.01887428378833839,
105
+ "grad_norm": 0.8125,
106
+ "learning_rate": 0.0003906,
107
+ "loss": 5.028667831420899,
108
+ "step": 280
109
+ },
110
+ {
111
+ "epoch": 0.020222446916076844,
112
+ "grad_norm": 0.75390625,
113
+ "learning_rate": 0.0004186,
114
+ "loss": 4.791546249389649,
115
+ "step": 300
116
+ },
117
+ {
118
+ "epoch": 0.021570610043815303,
119
+ "grad_norm": 0.875,
120
+ "learning_rate": 0.0004466,
121
+ "loss": 4.598950958251953,
122
+ "step": 320
123
+ },
124
+ {
125
+ "epoch": 0.022918773171553757,
126
+ "grad_norm": 0.8125,
127
+ "learning_rate": 0.00047460000000000004,
128
+ "loss": 4.41942024230957,
129
+ "step": 340
130
+ },
131
+ {
132
+ "epoch": 0.024266936299292215,
133
+ "grad_norm": 0.890625,
134
+ "learning_rate": 0.0005026,
135
+ "loss": 4.27253303527832,
136
+ "step": 360
137
+ },
138
+ {
139
+ "epoch": 0.02561509942703067,
140
+ "grad_norm": 1.2734375,
141
+ "learning_rate": 0.0005306,
142
+ "loss": 4.1283222198486325,
143
+ "step": 380
144
+ },
145
+ {
146
+ "epoch": 0.026963262554769128,
147
+ "grad_norm": 0.54296875,
148
+ "learning_rate": 0.0005586,
149
+ "loss": 3.9935020446777343,
150
+ "step": 400
151
+ },
152
+ {
153
+ "epoch": 0.028311425682507583,
154
+ "grad_norm": 0.546875,
155
+ "learning_rate": 0.0005866,
156
+ "loss": 3.885255813598633,
157
+ "step": 420
158
+ },
159
+ {
160
+ "epoch": 0.02965958881024604,
161
+ "grad_norm": 0.7890625,
162
+ "learning_rate": 0.0006146,
163
+ "loss": 3.821950912475586,
164
+ "step": 440
165
+ },
166
+ {
167
+ "epoch": 0.031007751937984496,
168
+ "grad_norm": 0.52734375,
169
+ "learning_rate": 0.0006426,
170
+ "loss": 3.7068336486816404,
171
+ "step": 460
172
+ },
173
+ {
174
+ "epoch": 0.032355915065722954,
175
+ "grad_norm": 0.6328125,
176
+ "learning_rate": 0.0006705999999999999,
177
+ "loss": 3.6370723724365233,
178
+ "step": 480
179
+ },
180
+ {
181
+ "epoch": 0.03370407819346141,
182
+ "grad_norm": 0.5625,
183
+ "learning_rate": 0.0006986,
184
+ "loss": 3.5550304412841798,
185
+ "step": 500
186
+ },
187
+ {
188
+ "epoch": 0.03505224132119986,
189
+ "grad_norm": 0.47265625,
190
+ "learning_rate": 0.0007,
191
+ "loss": 3.4874305725097656,
192
+ "step": 520
193
+ },
194
+ {
195
+ "epoch": 0.03640040444893832,
196
+ "grad_norm": 0.6328125,
197
+ "learning_rate": 0.0007,
198
+ "loss": 3.414617919921875,
199
+ "step": 540
200
+ },
201
+ {
202
+ "epoch": 0.03774856757667678,
203
+ "grad_norm": 0.478515625,
204
+ "learning_rate": 0.0007,
205
+ "loss": 3.3621952056884767,
206
+ "step": 560
207
+ },
208
+ {
209
+ "epoch": 0.03909673070441524,
210
+ "grad_norm": 0.478515625,
211
+ "learning_rate": 0.0007,
212
+ "loss": 3.301148223876953,
213
+ "step": 580
214
+ },
215
+ {
216
+ "epoch": 0.04044489383215369,
217
+ "grad_norm": 0.44140625,
218
+ "learning_rate": 0.0007,
219
+ "loss": 3.266124725341797,
220
+ "step": 600
221
+ },
222
+ {
223
+ "epoch": 0.04179305695989215,
224
+ "grad_norm": 0.58203125,
225
+ "learning_rate": 0.0007,
226
+ "loss": 3.1991281509399414,
227
+ "step": 620
228
+ },
229
+ {
230
+ "epoch": 0.043141220087630605,
231
+ "grad_norm": 0.51953125,
232
+ "learning_rate": 0.0007,
233
+ "loss": 3.175183868408203,
234
+ "step": 640
235
+ },
236
+ {
237
+ "epoch": 0.044489383215369056,
238
+ "grad_norm": 0.443359375,
239
+ "learning_rate": 0.0007,
240
+ "loss": 3.1185443878173826,
241
+ "step": 660
242
+ },
243
+ {
244
+ "epoch": 0.045837546343107514,
245
+ "grad_norm": 0.447265625,
246
+ "learning_rate": 0.0007,
247
+ "loss": 3.091288757324219,
248
+ "step": 680
249
+ },
250
+ {
251
+ "epoch": 0.04718570947084597,
252
+ "grad_norm": 0.400390625,
253
+ "learning_rate": 0.0007,
254
+ "loss": 3.0479175567626955,
255
+ "step": 700
256
+ },
257
+ {
258
+ "epoch": 0.04853387259858443,
259
+ "grad_norm": 0.39453125,
260
+ "learning_rate": 0.0007,
261
+ "loss": 3.012004280090332,
262
+ "step": 720
263
+ },
264
+ {
265
+ "epoch": 0.04988203572632288,
266
+ "grad_norm": 0.40625,
267
+ "learning_rate": 0.0007,
268
+ "loss": 2.973306655883789,
269
+ "step": 740
270
+ },
271
+ {
272
+ "epoch": 0.05123019885406134,
273
+ "grad_norm": 0.41015625,
274
+ "learning_rate": 0.0007,
275
+ "loss": 2.9569366455078123,
276
+ "step": 760
277
+ },
278
+ {
279
+ "epoch": 0.0525783619817998,
280
+ "grad_norm": 0.427734375,
281
+ "learning_rate": 0.0007,
282
+ "loss": 2.959245491027832,
283
+ "step": 780
284
+ },
285
+ {
286
+ "epoch": 0.053926525109538256,
287
+ "grad_norm": 0.4296875,
288
+ "learning_rate": 0.0007,
289
+ "loss": 2.9073596954345704,
290
+ "step": 800
291
+ },
292
+ {
293
+ "epoch": 0.05527468823727671,
294
+ "grad_norm": 0.435546875,
295
+ "learning_rate": 0.0007,
296
+ "loss": 2.894051933288574,
297
+ "step": 820
298
+ },
299
+ {
300
+ "epoch": 0.056622851365015166,
301
+ "grad_norm": 0.4296875,
302
+ "learning_rate": 0.0007,
303
+ "loss": 2.86330623626709,
304
+ "step": 840
305
+ },
306
+ {
307
+ "epoch": 0.057971014492753624,
308
+ "grad_norm": 0.388671875,
309
+ "learning_rate": 0.0007,
310
+ "loss": 2.81943302154541,
311
+ "step": 860
312
+ },
313
+ {
314
+ "epoch": 0.05931917762049208,
315
+ "grad_norm": 0.384765625,
316
+ "learning_rate": 0.0007,
317
+ "loss": 2.7944929122924806,
318
+ "step": 880
319
+ },
320
+ {
321
+ "epoch": 0.06066734074823053,
322
+ "grad_norm": 0.40234375,
323
+ "learning_rate": 0.0007,
324
+ "loss": 2.779962730407715,
325
+ "step": 900
326
+ },
327
+ {
328
+ "epoch": 0.06201550387596899,
329
+ "grad_norm": 0.416015625,
330
+ "learning_rate": 0.0007,
331
+ "loss": 2.751751708984375,
332
+ "step": 920
333
+ },
334
+ {
335
+ "epoch": 0.06336366700370745,
336
+ "grad_norm": 0.447265625,
337
+ "learning_rate": 0.0007,
338
+ "loss": 2.7392576217651365,
339
+ "step": 940
340
+ },
341
+ {
342
+ "epoch": 0.06471183013144591,
343
+ "grad_norm": 0.44140625,
344
+ "learning_rate": 0.0007,
345
+ "loss": 2.7111114501953124,
346
+ "step": 960
347
+ },
348
+ {
349
+ "epoch": 0.06605999325918437,
350
+ "grad_norm": 0.39453125,
351
+ "learning_rate": 0.0007,
352
+ "loss": 2.692020797729492,
353
+ "step": 980
354
+ },
355
+ {
356
+ "epoch": 0.06740815638692282,
357
+ "grad_norm": 0.36328125,
358
+ "learning_rate": 0.0007,
359
+ "loss": 2.691050910949707,
360
+ "step": 1000
361
+ }
362
+ ],
363
+ "logging_steps": 20,
364
+ "max_steps": 2500,
365
+ "num_input_tokens_seen": 0,
366
+ "num_train_epochs": 1,
367
+ "save_steps": 100,
368
+ "stateful_callbacks": {
369
+ "TrainerControl": {
370
+ "args": {
371
+ "should_epoch_stop": false,
372
+ "should_evaluate": false,
373
+ "should_log": false,
374
+ "should_save": true,
375
+ "should_training_stop": false
376
+ },
377
+ "attributes": {}
378
+ }
379
+ },
380
+ "total_flos": 3226283802624000.0,
381
+ "train_batch_size": 64,
382
+ "trial_name": null,
383
+ "trial_params": null
384
+ }
zain/Activation/out/glu-situglu-100L_run/checkpoint-1000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26b99898fe48ff0e75b1f0139845f22400023db32e87c8fca3404fda1b772936
3
+ size 4920
zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "situglu",
3
+ "architectures": [
4
+ "TinyLlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 2,
11
+ "head_dim": 32,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 128,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 256,
16
+ "max_position_embeddings": 512,
17
+ "mlp_bias": false,
18
+ "mlp_type": "glu",
19
+ "model_type": "tiny_llama",
20
+ "num_attention_heads": 4,
21
+ "num_hidden_layers": 100,
22
+ "num_key_value_heads": 4,
23
+ "pad_token_id": 0,
24
+ "pretraining_tp": 1,
25
+ "rms_norm_eps": 1e-06,
26
+ "rope_parameters": {
27
+ "rope_theta": 10000.0,
28
+ "rope_type": "default"
29
+ },
30
+ "tie_word_embeddings": true,
31
+ "tokenizer_name": "w-ahmad/tiny-stories-tokenizer",
32
+ "transformers_version": "5.16.0.dev0",
33
+ "use_cache": false,
34
+ "vocab_size": 4096,
35
+ "waleed_beta": 10.0
36
+ }
zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a9c7adeda4e8598473854f66d131aad801585278c3d312d86bd467facf00505f
3
+ size 33967272
zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ea77ebf2ec0eb8ff20f9f42a4cf67f6eb9b7f584fa772b06f4ce177c74ae20d8
3
+ size 68504996
zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9d9cd6a0487226e5bd30d1846894c82af483733ab4381b75bae9c0745e05d405
3
+ size 14244
zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:88121baf7de31688d97eca5c32cb058d585673d52f936d32dcd04e8b1903dcac
3
+ size 1064
zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/tokenizer_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|endoftext|>",
5
+ "eos_token": "<|endoftext|>",
6
+ "errors": "replace",
7
+ "is_local": false,
8
+ "local_files_only": false,
9
+ "model_max_length": 1024,
10
+ "pad_token": "<|endoftext|>",
11
+ "tokenizer_class": "GPT2Tokenizer",
12
+ "unk_token": "<|endoftext|>"
13
+ }
zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/trainer_state.json ADDED
@@ -0,0 +1,419 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.0741489720256151,
6
+ "eval_steps": 2498,
7
+ "global_step": 1100,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0013481631277384564,
14
+ "grad_norm": 2.078125,
15
+ "learning_rate": 2.66e-05,
16
+ "loss": 8.299653625488281,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.3515625,
22
+ "learning_rate": 5.46e-05,
23
+ "loss": 8.04769058227539,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 1.2734375,
29
+ "learning_rate": 8.259999999999999e-05,
30
+ "loss": 7.8379150390625,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 1.265625,
36
+ "learning_rate": 0.0001106,
37
+ "loss": 7.572837066650391,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 1.1953125,
43
+ "learning_rate": 0.0001386,
44
+ "loss": 7.233182525634765,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 1.1328125,
50
+ "learning_rate": 0.00016659999999999998,
51
+ "loss": 6.860689544677735,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 0.94921875,
57
+ "learning_rate": 0.00019460000000000001,
58
+ "loss": 6.517751312255859,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 0.8046875,
64
+ "learning_rate": 0.0002226,
65
+ "loss": 6.239968109130859,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 0.6640625,
71
+ "learning_rate": 0.00025059999999999997,
72
+ "loss": 6.041232299804688,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 0.5234375,
78
+ "learning_rate": 0.0002786,
79
+ "loss": 5.863973999023438,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.01482979440512302,
84
+ "grad_norm": 1.375,
85
+ "learning_rate": 0.00030659999999999997,
86
+ "loss": 5.692941665649414,
87
+ "step": 220
88
+ },
89
+ {
90
+ "epoch": 0.016177957532861477,
91
+ "grad_norm": 0.875,
92
+ "learning_rate": 0.0003346,
93
+ "loss": 5.484311294555664,
94
+ "step": 240
95
+ },
96
+ {
97
+ "epoch": 0.01752612066059993,
98
+ "grad_norm": 1.25,
99
+ "learning_rate": 0.00036260000000000003,
100
+ "loss": 5.25721549987793,
101
+ "step": 260
102
+ },
103
+ {
104
+ "epoch": 0.01887428378833839,
105
+ "grad_norm": 0.8125,
106
+ "learning_rate": 0.0003906,
107
+ "loss": 5.028667831420899,
108
+ "step": 280
109
+ },
110
+ {
111
+ "epoch": 0.020222446916076844,
112
+ "grad_norm": 0.75390625,
113
+ "learning_rate": 0.0004186,
114
+ "loss": 4.791546249389649,
115
+ "step": 300
116
+ },
117
+ {
118
+ "epoch": 0.021570610043815303,
119
+ "grad_norm": 0.875,
120
+ "learning_rate": 0.0004466,
121
+ "loss": 4.598950958251953,
122
+ "step": 320
123
+ },
124
+ {
125
+ "epoch": 0.022918773171553757,
126
+ "grad_norm": 0.8125,
127
+ "learning_rate": 0.00047460000000000004,
128
+ "loss": 4.41942024230957,
129
+ "step": 340
130
+ },
131
+ {
132
+ "epoch": 0.024266936299292215,
133
+ "grad_norm": 0.890625,
134
+ "learning_rate": 0.0005026,
135
+ "loss": 4.27253303527832,
136
+ "step": 360
137
+ },
138
+ {
139
+ "epoch": 0.02561509942703067,
140
+ "grad_norm": 1.2734375,
141
+ "learning_rate": 0.0005306,
142
+ "loss": 4.1283222198486325,
143
+ "step": 380
144
+ },
145
+ {
146
+ "epoch": 0.026963262554769128,
147
+ "grad_norm": 0.54296875,
148
+ "learning_rate": 0.0005586,
149
+ "loss": 3.9935020446777343,
150
+ "step": 400
151
+ },
152
+ {
153
+ "epoch": 0.028311425682507583,
154
+ "grad_norm": 0.546875,
155
+ "learning_rate": 0.0005866,
156
+ "loss": 3.885255813598633,
157
+ "step": 420
158
+ },
159
+ {
160
+ "epoch": 0.02965958881024604,
161
+ "grad_norm": 0.7890625,
162
+ "learning_rate": 0.0006146,
163
+ "loss": 3.821950912475586,
164
+ "step": 440
165
+ },
166
+ {
167
+ "epoch": 0.031007751937984496,
168
+ "grad_norm": 0.52734375,
169
+ "learning_rate": 0.0006426,
170
+ "loss": 3.7068336486816404,
171
+ "step": 460
172
+ },
173
+ {
174
+ "epoch": 0.032355915065722954,
175
+ "grad_norm": 0.6328125,
176
+ "learning_rate": 0.0006705999999999999,
177
+ "loss": 3.6370723724365233,
178
+ "step": 480
179
+ },
180
+ {
181
+ "epoch": 0.03370407819346141,
182
+ "grad_norm": 0.5625,
183
+ "learning_rate": 0.0006986,
184
+ "loss": 3.5550304412841798,
185
+ "step": 500
186
+ },
187
+ {
188
+ "epoch": 0.03505224132119986,
189
+ "grad_norm": 0.47265625,
190
+ "learning_rate": 0.0007,
191
+ "loss": 3.4874305725097656,
192
+ "step": 520
193
+ },
194
+ {
195
+ "epoch": 0.03640040444893832,
196
+ "grad_norm": 0.6328125,
197
+ "learning_rate": 0.0007,
198
+ "loss": 3.414617919921875,
199
+ "step": 540
200
+ },
201
+ {
202
+ "epoch": 0.03774856757667678,
203
+ "grad_norm": 0.478515625,
204
+ "learning_rate": 0.0007,
205
+ "loss": 3.3621952056884767,
206
+ "step": 560
207
+ },
208
+ {
209
+ "epoch": 0.03909673070441524,
210
+ "grad_norm": 0.478515625,
211
+ "learning_rate": 0.0007,
212
+ "loss": 3.301148223876953,
213
+ "step": 580
214
+ },
215
+ {
216
+ "epoch": 0.04044489383215369,
217
+ "grad_norm": 0.44140625,
218
+ "learning_rate": 0.0007,
219
+ "loss": 3.266124725341797,
220
+ "step": 600
221
+ },
222
+ {
223
+ "epoch": 0.04179305695989215,
224
+ "grad_norm": 0.58203125,
225
+ "learning_rate": 0.0007,
226
+ "loss": 3.1991281509399414,
227
+ "step": 620
228
+ },
229
+ {
230
+ "epoch": 0.043141220087630605,
231
+ "grad_norm": 0.51953125,
232
+ "learning_rate": 0.0007,
233
+ "loss": 3.175183868408203,
234
+ "step": 640
235
+ },
236
+ {
237
+ "epoch": 0.044489383215369056,
238
+ "grad_norm": 0.443359375,
239
+ "learning_rate": 0.0007,
240
+ "loss": 3.1185443878173826,
241
+ "step": 660
242
+ },
243
+ {
244
+ "epoch": 0.045837546343107514,
245
+ "grad_norm": 0.447265625,
246
+ "learning_rate": 0.0007,
247
+ "loss": 3.091288757324219,
248
+ "step": 680
249
+ },
250
+ {
251
+ "epoch": 0.04718570947084597,
252
+ "grad_norm": 0.400390625,
253
+ "learning_rate": 0.0007,
254
+ "loss": 3.0479175567626955,
255
+ "step": 700
256
+ },
257
+ {
258
+ "epoch": 0.04853387259858443,
259
+ "grad_norm": 0.39453125,
260
+ "learning_rate": 0.0007,
261
+ "loss": 3.012004280090332,
262
+ "step": 720
263
+ },
264
+ {
265
+ "epoch": 0.04988203572632288,
266
+ "grad_norm": 0.40625,
267
+ "learning_rate": 0.0007,
268
+ "loss": 2.973306655883789,
269
+ "step": 740
270
+ },
271
+ {
272
+ "epoch": 0.05123019885406134,
273
+ "grad_norm": 0.41015625,
274
+ "learning_rate": 0.0007,
275
+ "loss": 2.9569366455078123,
276
+ "step": 760
277
+ },
278
+ {
279
+ "epoch": 0.0525783619817998,
280
+ "grad_norm": 0.427734375,
281
+ "learning_rate": 0.0007,
282
+ "loss": 2.959245491027832,
283
+ "step": 780
284
+ },
285
+ {
286
+ "epoch": 0.053926525109538256,
287
+ "grad_norm": 0.4296875,
288
+ "learning_rate": 0.0007,
289
+ "loss": 2.9073596954345704,
290
+ "step": 800
291
+ },
292
+ {
293
+ "epoch": 0.05527468823727671,
294
+ "grad_norm": 0.435546875,
295
+ "learning_rate": 0.0007,
296
+ "loss": 2.894051933288574,
297
+ "step": 820
298
+ },
299
+ {
300
+ "epoch": 0.056622851365015166,
301
+ "grad_norm": 0.4296875,
302
+ "learning_rate": 0.0007,
303
+ "loss": 2.86330623626709,
304
+ "step": 840
305
+ },
306
+ {
307
+ "epoch": 0.057971014492753624,
308
+ "grad_norm": 0.388671875,
309
+ "learning_rate": 0.0007,
310
+ "loss": 2.81943302154541,
311
+ "step": 860
312
+ },
313
+ {
314
+ "epoch": 0.05931917762049208,
315
+ "grad_norm": 0.384765625,
316
+ "learning_rate": 0.0007,
317
+ "loss": 2.7944929122924806,
318
+ "step": 880
319
+ },
320
+ {
321
+ "epoch": 0.06066734074823053,
322
+ "grad_norm": 0.40234375,
323
+ "learning_rate": 0.0007,
324
+ "loss": 2.779962730407715,
325
+ "step": 900
326
+ },
327
+ {
328
+ "epoch": 0.06201550387596899,
329
+ "grad_norm": 0.416015625,
330
+ "learning_rate": 0.0007,
331
+ "loss": 2.751751708984375,
332
+ "step": 920
333
+ },
334
+ {
335
+ "epoch": 0.06336366700370745,
336
+ "grad_norm": 0.447265625,
337
+ "learning_rate": 0.0007,
338
+ "loss": 2.7392576217651365,
339
+ "step": 940
340
+ },
341
+ {
342
+ "epoch": 0.06471183013144591,
343
+ "grad_norm": 0.44140625,
344
+ "learning_rate": 0.0007,
345
+ "loss": 2.7111114501953124,
346
+ "step": 960
347
+ },
348
+ {
349
+ "epoch": 0.06605999325918437,
350
+ "grad_norm": 0.39453125,
351
+ "learning_rate": 0.0007,
352
+ "loss": 2.692020797729492,
353
+ "step": 980
354
+ },
355
+ {
356
+ "epoch": 0.06740815638692282,
357
+ "grad_norm": 0.36328125,
358
+ "learning_rate": 0.0007,
359
+ "loss": 2.691050910949707,
360
+ "step": 1000
361
+ },
362
+ {
363
+ "epoch": 0.06875631951466127,
364
+ "grad_norm": 0.37109375,
365
+ "learning_rate": 0.0007,
366
+ "loss": 2.6589706420898436,
367
+ "step": 1020
368
+ },
369
+ {
370
+ "epoch": 0.07010448264239973,
371
+ "grad_norm": 0.40625,
372
+ "learning_rate": 0.0007,
373
+ "loss": 2.643506622314453,
374
+ "step": 1040
375
+ },
376
+ {
377
+ "epoch": 0.07145264577013818,
378
+ "grad_norm": 0.390625,
379
+ "learning_rate": 0.0007,
380
+ "loss": 2.629362869262695,
381
+ "step": 1060
382
+ },
383
+ {
384
+ "epoch": 0.07280080889787664,
385
+ "grad_norm": 0.37890625,
386
+ "learning_rate": 0.0007,
387
+ "loss": 2.606895637512207,
388
+ "step": 1080
389
+ },
390
+ {
391
+ "epoch": 0.0741489720256151,
392
+ "grad_norm": 0.412109375,
393
+ "learning_rate": 0.0007,
394
+ "loss": 2.6003129959106444,
395
+ "step": 1100
396
+ }
397
+ ],
398
+ "logging_steps": 20,
399
+ "max_steps": 2500,
400
+ "num_input_tokens_seen": 0,
401
+ "num_train_epochs": 1,
402
+ "save_steps": 100,
403
+ "stateful_callbacks": {
404
+ "TrainerControl": {
405
+ "args": {
406
+ "should_epoch_stop": false,
407
+ "should_evaluate": false,
408
+ "should_log": false,
409
+ "should_save": true,
410
+ "should_training_stop": false
411
+ },
412
+ "attributes": {}
413
+ }
414
+ },
415
+ "total_flos": 3548912182886400.0,
416
+ "train_batch_size": 64,
417
+ "trial_name": null,
418
+ "trial_params": null
419
+ }
zain/Activation/out/glu-situglu-100L_run/checkpoint-1100/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26b99898fe48ff0e75b1f0139845f22400023db32e87c8fca3404fda1b772936
3
+ size 4920
zain/Activation/out/glu-situglu-100L_run/checkpoint-1200/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "situglu",
3
+ "architectures": [
4
+ "TinyLlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 2,
11
+ "head_dim": 32,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 128,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 256,
16
+ "max_position_embeddings": 512,
17
+ "mlp_bias": false,
18
+ "mlp_type": "glu",
19
+ "model_type": "tiny_llama",
20
+ "num_attention_heads": 4,
21
+ "num_hidden_layers": 100,
22
+ "num_key_value_heads": 4,
23
+ "pad_token_id": 0,
24
+ "pretraining_tp": 1,
25
+ "rms_norm_eps": 1e-06,
26
+ "rope_parameters": {
27
+ "rope_theta": 10000.0,
28
+ "rope_type": "default"
29
+ },
30
+ "tie_word_embeddings": true,
31
+ "tokenizer_name": "w-ahmad/tiny-stories-tokenizer",
32
+ "transformers_version": "5.16.0.dev0",
33
+ "use_cache": false,
34
+ "vocab_size": 4096,
35
+ "waleed_beta": 10.0
36
+ }
zain/Activation/out/glu-situglu-100L_run/checkpoint-1200/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be417a09a0876787811cb9093fcd79e61308654aad1cfc4b94421cc66a43c22c
3
+ size 33967272
zain/Activation/out/glu-situglu-100L_run/checkpoint-1200/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a18641cff33dafe21c59ba3cdc9b1785bea2758245e2af54e6aa8ca8a26a23a
3
+ size 68504996
zain/Activation/out/glu-situglu-100L_run/checkpoint-1200/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9d9cd6a0487226e5bd30d1846894c82af483733ab4381b75bae9c0745e05d405
3
+ size 14244
zain/Activation/out/glu-situglu-100L_run/checkpoint-1200/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8543afc9166865219cadbba29db865d839ec4102a4f9ddeffc962b9fdba3b028
3
+ size 1064
zain/Activation/out/glu-situglu-100L_run/checkpoint-1200/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-situglu-100L_run/checkpoint-1200/tokenizer_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|endoftext|>",
5
+ "eos_token": "<|endoftext|>",
6
+ "errors": "replace",
7
+ "is_local": false,
8
+ "local_files_only": false,
9
+ "model_max_length": 1024,
10
+ "pad_token": "<|endoftext|>",
11
+ "tokenizer_class": "GPT2Tokenizer",
12
+ "unk_token": "<|endoftext|>"
13
+ }