w-ahmad commited on
Commit
4e9d933
·
verified ·
1 Parent(s): 8d91c6a

Auto upload zain 2026-08-12T23:10:46.993782 (part 2)

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. zain/Activation/out/glu-waleed-100L_run/checkpoint-2000/config.json +36 -0
  2. zain/Activation/out/glu-waleed-100L_run/checkpoint-2000/model.safetensors +3 -0
  3. zain/Activation/out/glu-waleed-100L_run/checkpoint-2000/optimizer.pt +3 -0
  4. zain/Activation/out/glu-waleed-100L_run/checkpoint-2000/rng_state.pth +3 -0
  5. zain/Activation/out/glu-waleed-100L_run/checkpoint-2000/scheduler.pt +3 -0
  6. zain/Activation/out/glu-waleed-100L_run/checkpoint-2000/tokenizer.json +0 -0
  7. zain/Activation/out/glu-waleed-100L_run/checkpoint-2000/tokenizer_config.json +13 -0
  8. zain/Activation/out/glu-waleed-100L_run/checkpoint-2000/trainer_state.json +734 -0
  9. zain/Activation/out/glu-waleed-100L_run/checkpoint-2000/training_args.bin +3 -0
  10. zain/Activation/out/glu-waleed-100L_run/checkpoint-2100/config.json +36 -0
  11. zain/Activation/out/glu-waleed-100L_run/checkpoint-2100/model.safetensors +3 -0
  12. zain/Activation/out/glu-waleed-100L_run/checkpoint-2100/optimizer.pt +3 -0
  13. zain/Activation/out/glu-waleed-100L_run/checkpoint-2100/rng_state.pth +3 -0
  14. zain/Activation/out/glu-waleed-100L_run/checkpoint-2100/scheduler.pt +3 -0
  15. zain/Activation/out/glu-waleed-100L_run/checkpoint-2100/tokenizer.json +0 -0
  16. zain/Activation/out/glu-waleed-100L_run/checkpoint-2100/tokenizer_config.json +13 -0
  17. zain/Activation/out/glu-waleed-100L_run/checkpoint-2100/trainer_state.json +769 -0
  18. zain/Activation/out/glu-waleed-100L_run/checkpoint-2100/training_args.bin +3 -0
  19. zain/Activation/out/glu-waleed-100L_run/checkpoint-2200/config.json +36 -0
  20. zain/Activation/out/glu-waleed-100L_run/checkpoint-2200/model.safetensors +3 -0
  21. zain/Activation/out/glu-waleed-100L_run/checkpoint-2200/optimizer.pt +3 -0
  22. zain/Activation/out/glu-waleed-100L_run/checkpoint-2200/rng_state.pth +3 -0
  23. zain/Activation/out/glu-waleed-100L_run/checkpoint-2200/scheduler.pt +3 -0
  24. zain/Activation/out/glu-waleed-100L_run/checkpoint-2200/tokenizer.json +0 -0
  25. zain/Activation/out/glu-waleed-100L_run/checkpoint-2200/tokenizer_config.json +13 -0
  26. zain/Activation/out/glu-waleed-100L_run/checkpoint-2200/trainer_state.json +804 -0
  27. zain/Activation/out/glu-waleed-100L_run/checkpoint-2200/training_args.bin +3 -0
  28. zain/Activation/out/glu-waleed-100L_run/checkpoint-2300/config.json +36 -0
  29. zain/Activation/out/glu-waleed-100L_run/checkpoint-2300/model.safetensors +3 -0
  30. zain/Activation/out/glu-waleed-100L_run/checkpoint-2300/optimizer.pt +3 -0
  31. zain/Activation/out/glu-waleed-100L_run/checkpoint-2300/rng_state.pth +3 -0
  32. zain/Activation/out/glu-waleed-100L_run/checkpoint-2300/scheduler.pt +3 -0
  33. zain/Activation/out/glu-waleed-100L_run/checkpoint-2300/tokenizer.json +0 -0
  34. zain/Activation/out/glu-waleed-100L_run/checkpoint-2300/tokenizer_config.json +13 -0
  35. zain/Activation/out/glu-waleed-100L_run/checkpoint-2300/trainer_state.json +839 -0
  36. zain/Activation/out/glu-waleed-100L_run/checkpoint-2300/training_args.bin +3 -0
  37. zain/Activation/out/glu-waleed-100L_run/checkpoint-2400/config.json +36 -0
  38. zain/Activation/out/glu-waleed-100L_run/checkpoint-2400/model.safetensors +3 -0
  39. zain/Activation/out/glu-waleed-100L_run/checkpoint-2400/optimizer.pt +3 -0
  40. zain/Activation/out/glu-waleed-100L_run/checkpoint-2400/rng_state.pth +3 -0
  41. zain/Activation/out/glu-waleed-100L_run/checkpoint-2400/scheduler.pt +3 -0
  42. zain/Activation/out/glu-waleed-100L_run/checkpoint-2400/tokenizer.json +0 -0
  43. zain/Activation/out/glu-waleed-100L_run/checkpoint-2400/tokenizer_config.json +13 -0
  44. zain/Activation/out/glu-waleed-100L_run/checkpoint-2400/trainer_state.json +874 -0
  45. zain/Activation/out/glu-waleed-100L_run/checkpoint-2400/training_args.bin +3 -0
  46. zain/Activation/out/glu-waleed-100L_run/checkpoint-2500/config.json +36 -0
  47. zain/Activation/out/glu-waleed-100L_run/checkpoint-2500/model.safetensors +3 -0
  48. zain/Activation/out/glu-waleed-100L_run/checkpoint-2500/optimizer.pt +3 -0
  49. zain/Activation/out/glu-waleed-100L_run/checkpoint-2500/rng_state.pth +3 -0
  50. zain/Activation/out/glu-waleed-100L_run/checkpoint-2500/scheduler.pt +3 -0
zain/Activation/out/glu-waleed-100L_run/checkpoint-2000/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "waleed",
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-waleed-100L_run/checkpoint-2000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:23114f1be9620427155100ae8c09a716a097ee0b40a63944c4f73c21b9a7c452
3
+ size 33967272
zain/Activation/out/glu-waleed-100L_run/checkpoint-2000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a9ab733d35674310e93dd5b6407429742b97bf9532576e80d55f1a2d4a65c109
3
+ size 68504996
zain/Activation/out/glu-waleed-100L_run/checkpoint-2000/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-waleed-100L_run/checkpoint-2000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eb897171d6314e0d5ff47840624a03ce1078f22b8f8c3062289ef88ef22a406e
3
+ size 1064
zain/Activation/out/glu-waleed-100L_run/checkpoint-2000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-waleed-100L_run/checkpoint-2000/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-waleed-100L_run/checkpoint-2000/trainer_state.json ADDED
@@ -0,0 +1,734 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.13481631277384565,
6
+ "eval_steps": 2498,
7
+ "global_step": 2000,
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": 2.125,
15
+ "learning_rate": 2.66e-05,
16
+ "loss": 8.299404907226563,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.3125,
22
+ "learning_rate": 5.46e-05,
23
+ "loss": 8.036402130126953,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 1.2734375,
29
+ "learning_rate": 8.259999999999999e-05,
30
+ "loss": 7.834580993652343,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 1.265625,
36
+ "learning_rate": 0.0001106,
37
+ "loss": 7.5719047546386715,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 1.1953125,
43
+ "learning_rate": 0.0001386,
44
+ "loss": 7.237163543701172,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 1.1484375,
50
+ "learning_rate": 0.00016659999999999998,
51
+ "loss": 6.864958190917969,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 1.015625,
57
+ "learning_rate": 0.00019460000000000001,
58
+ "loss": 6.509477996826172,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 1.0234375,
64
+ "learning_rate": 0.0002226,
65
+ "loss": 6.204301452636718,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 1.0859375,
71
+ "learning_rate": 0.00025059999999999997,
72
+ "loss": 5.940593719482422,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 1.2109375,
78
+ "learning_rate": 0.0002786,
79
+ "loss": 5.676498794555664,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.01482979440512302,
84
+ "grad_norm": 0.8046875,
85
+ "learning_rate": 0.00030659999999999997,
86
+ "loss": 5.412863540649414,
87
+ "step": 220
88
+ },
89
+ {
90
+ "epoch": 0.016177957532861477,
91
+ "grad_norm": 0.71875,
92
+ "learning_rate": 0.0003346,
93
+ "loss": 5.1301216125488285,
94
+ "step": 240
95
+ },
96
+ {
97
+ "epoch": 0.01752612066059993,
98
+ "grad_norm": 1.046875,
99
+ "learning_rate": 0.00036260000000000003,
100
+ "loss": 4.88237190246582,
101
+ "step": 260
102
+ },
103
+ {
104
+ "epoch": 0.01887428378833839,
105
+ "grad_norm": 0.75,
106
+ "learning_rate": 0.0003906,
107
+ "loss": 4.643666076660156,
108
+ "step": 280
109
+ },
110
+ {
111
+ "epoch": 0.020222446916076844,
112
+ "grad_norm": 0.83984375,
113
+ "learning_rate": 0.0004186,
114
+ "loss": 4.441861343383789,
115
+ "step": 300
116
+ },
117
+ {
118
+ "epoch": 0.021570610043815303,
119
+ "grad_norm": 0.8125,
120
+ "learning_rate": 0.0004466,
121
+ "loss": 4.25420036315918,
122
+ "step": 320
123
+ },
124
+ {
125
+ "epoch": 0.022918773171553757,
126
+ "grad_norm": 0.62109375,
127
+ "learning_rate": 0.00047460000000000004,
128
+ "loss": 4.097484970092774,
129
+ "step": 340
130
+ },
131
+ {
132
+ "epoch": 0.024266936299292215,
133
+ "grad_norm": 0.6015625,
134
+ "learning_rate": 0.0005026,
135
+ "loss": 3.9764480590820312,
136
+ "step": 360
137
+ },
138
+ {
139
+ "epoch": 0.02561509942703067,
140
+ "grad_norm": 0.9140625,
141
+ "learning_rate": 0.0005306,
142
+ "loss": 3.854802703857422,
143
+ "step": 380
144
+ },
145
+ {
146
+ "epoch": 0.026963262554769128,
147
+ "grad_norm": 0.56640625,
148
+ "learning_rate": 0.0005586,
149
+ "loss": 3.7329875946044924,
150
+ "step": 400
151
+ },
152
+ {
153
+ "epoch": 0.028311425682507583,
154
+ "grad_norm": 0.53515625,
155
+ "learning_rate": 0.0005866,
156
+ "loss": 3.6484100341796877,
157
+ "step": 420
158
+ },
159
+ {
160
+ "epoch": 0.02965958881024604,
161
+ "grad_norm": 0.5390625,
162
+ "learning_rate": 0.0006146,
163
+ "loss": 3.595387268066406,
164
+ "step": 440
165
+ },
166
+ {
167
+ "epoch": 0.031007751937984496,
168
+ "grad_norm": 0.55859375,
169
+ "learning_rate": 0.0006426,
170
+ "loss": 3.498302459716797,
171
+ "step": 460
172
+ },
173
+ {
174
+ "epoch": 0.032355915065722954,
175
+ "grad_norm": 0.5859375,
176
+ "learning_rate": 0.0006705999999999999,
177
+ "loss": 3.4313674926757813,
178
+ "step": 480
179
+ },
180
+ {
181
+ "epoch": 0.03370407819346141,
182
+ "grad_norm": 0.61328125,
183
+ "learning_rate": 0.0006986,
184
+ "loss": 3.3613121032714846,
185
+ "step": 500
186
+ },
187
+ {
188
+ "epoch": 0.03505224132119986,
189
+ "grad_norm": 0.474609375,
190
+ "learning_rate": 0.0007,
191
+ "loss": 3.3039566040039063,
192
+ "step": 520
193
+ },
194
+ {
195
+ "epoch": 0.03640040444893832,
196
+ "grad_norm": 0.578125,
197
+ "learning_rate": 0.0007,
198
+ "loss": 3.2328567504882812,
199
+ "step": 540
200
+ },
201
+ {
202
+ "epoch": 0.03774856757667678,
203
+ "grad_norm": 0.64453125,
204
+ "learning_rate": 0.0007,
205
+ "loss": 3.187940788269043,
206
+ "step": 560
207
+ },
208
+ {
209
+ "epoch": 0.03909673070441524,
210
+ "grad_norm": 0.462890625,
211
+ "learning_rate": 0.0007,
212
+ "loss": 3.133651542663574,
213
+ "step": 580
214
+ },
215
+ {
216
+ "epoch": 0.04044489383215369,
217
+ "grad_norm": 0.5,
218
+ "learning_rate": 0.0007,
219
+ "loss": 3.1042255401611327,
220
+ "step": 600
221
+ },
222
+ {
223
+ "epoch": 0.04179305695989215,
224
+ "grad_norm": 0.578125,
225
+ "learning_rate": 0.0007,
226
+ "loss": 3.04781551361084,
227
+ "step": 620
228
+ },
229
+ {
230
+ "epoch": 0.043141220087630605,
231
+ "grad_norm": 0.453125,
232
+ "learning_rate": 0.0007,
233
+ "loss": 3.0298320770263674,
234
+ "step": 640
235
+ },
236
+ {
237
+ "epoch": 0.044489383215369056,
238
+ "grad_norm": 0.46875,
239
+ "learning_rate": 0.0007,
240
+ "loss": 2.9790300369262694,
241
+ "step": 660
242
+ },
243
+ {
244
+ "epoch": 0.045837546343107514,
245
+ "grad_norm": 0.46875,
246
+ "learning_rate": 0.0007,
247
+ "loss": 2.9566295623779295,
248
+ "step": 680
249
+ },
250
+ {
251
+ "epoch": 0.04718570947084597,
252
+ "grad_norm": 0.439453125,
253
+ "learning_rate": 0.0007,
254
+ "loss": 2.916979217529297,
255
+ "step": 700
256
+ },
257
+ {
258
+ "epoch": 0.04853387259858443,
259
+ "grad_norm": 0.41796875,
260
+ "learning_rate": 0.0007,
261
+ "loss": 2.8888761520385744,
262
+ "step": 720
263
+ },
264
+ {
265
+ "epoch": 0.04988203572632288,
266
+ "grad_norm": 0.4765625,
267
+ "learning_rate": 0.0007,
268
+ "loss": 2.855138969421387,
269
+ "step": 740
270
+ },
271
+ {
272
+ "epoch": 0.05123019885406134,
273
+ "grad_norm": 0.46875,
274
+ "learning_rate": 0.0007,
275
+ "loss": 2.841838836669922,
276
+ "step": 760
277
+ },
278
+ {
279
+ "epoch": 0.0525783619817998,
280
+ "grad_norm": 0.43359375,
281
+ "learning_rate": 0.0007,
282
+ "loss": 2.8384235382080076,
283
+ "step": 780
284
+ },
285
+ {
286
+ "epoch": 0.053926525109538256,
287
+ "grad_norm": 0.447265625,
288
+ "learning_rate": 0.0007,
289
+ "loss": 2.7947757720947264,
290
+ "step": 800
291
+ },
292
+ {
293
+ "epoch": 0.05527468823727671,
294
+ "grad_norm": 0.478515625,
295
+ "learning_rate": 0.0007,
296
+ "loss": 2.7847801208496095,
297
+ "step": 820
298
+ },
299
+ {
300
+ "epoch": 0.056622851365015166,
301
+ "grad_norm": 0.44140625,
302
+ "learning_rate": 0.0007,
303
+ "loss": 2.7601039886474608,
304
+ "step": 840
305
+ },
306
+ {
307
+ "epoch": 0.057971014492753624,
308
+ "grad_norm": 0.484375,
309
+ "learning_rate": 0.0007,
310
+ "loss": 2.716906929016113,
311
+ "step": 860
312
+ },
313
+ {
314
+ "epoch": 0.05931917762049208,
315
+ "grad_norm": 0.439453125,
316
+ "learning_rate": 0.0007,
317
+ "loss": 2.6968852996826174,
318
+ "step": 880
319
+ },
320
+ {
321
+ "epoch": 0.06066734074823053,
322
+ "grad_norm": 0.447265625,
323
+ "learning_rate": 0.0007,
324
+ "loss": 2.6847219467163086,
325
+ "step": 900
326
+ },
327
+ {
328
+ "epoch": 0.06201550387596899,
329
+ "grad_norm": 0.435546875,
330
+ "learning_rate": 0.0007,
331
+ "loss": 2.6599369049072266,
332
+ "step": 920
333
+ },
334
+ {
335
+ "epoch": 0.06336366700370745,
336
+ "grad_norm": 0.455078125,
337
+ "learning_rate": 0.0007,
338
+ "loss": 2.6465431213378907,
339
+ "step": 940
340
+ },
341
+ {
342
+ "epoch": 0.06471183013144591,
343
+ "grad_norm": 0.455078125,
344
+ "learning_rate": 0.0007,
345
+ "loss": 2.623439979553223,
346
+ "step": 960
347
+ },
348
+ {
349
+ "epoch": 0.06605999325918437,
350
+ "grad_norm": 0.462890625,
351
+ "learning_rate": 0.0007,
352
+ "loss": 2.609505271911621,
353
+ "step": 980
354
+ },
355
+ {
356
+ "epoch": 0.06740815638692282,
357
+ "grad_norm": 0.48828125,
358
+ "learning_rate": 0.0007,
359
+ "loss": 2.610812187194824,
360
+ "step": 1000
361
+ },
362
+ {
363
+ "epoch": 0.06875631951466127,
364
+ "grad_norm": 0.423828125,
365
+ "learning_rate": 0.0007,
366
+ "loss": 2.581887054443359,
367
+ "step": 1020
368
+ },
369
+ {
370
+ "epoch": 0.07010448264239973,
371
+ "grad_norm": 0.4453125,
372
+ "learning_rate": 0.0007,
373
+ "loss": 2.5679080963134764,
374
+ "step": 1040
375
+ },
376
+ {
377
+ "epoch": 0.07145264577013818,
378
+ "grad_norm": 0.447265625,
379
+ "learning_rate": 0.0007,
380
+ "loss": 2.55521240234375,
381
+ "step": 1060
382
+ },
383
+ {
384
+ "epoch": 0.07280080889787664,
385
+ "grad_norm": 0.4453125,
386
+ "learning_rate": 0.0007,
387
+ "loss": 2.536572074890137,
388
+ "step": 1080
389
+ },
390
+ {
391
+ "epoch": 0.0741489720256151,
392
+ "grad_norm": 0.431640625,
393
+ "learning_rate": 0.0007,
394
+ "loss": 2.5341920852661133,
395
+ "step": 1100
396
+ },
397
+ {
398
+ "epoch": 0.07549713515335356,
399
+ "grad_norm": 0.427734375,
400
+ "learning_rate": 0.0007,
401
+ "loss": 2.4994813919067385,
402
+ "step": 1120
403
+ },
404
+ {
405
+ "epoch": 0.07684529828109202,
406
+ "grad_norm": 0.421875,
407
+ "learning_rate": 0.0007,
408
+ "loss": 2.4886734008789064,
409
+ "step": 1140
410
+ },
411
+ {
412
+ "epoch": 0.07819346140883048,
413
+ "grad_norm": 0.419921875,
414
+ "learning_rate": 0.0007,
415
+ "loss": 2.4889175415039064,
416
+ "step": 1160
417
+ },
418
+ {
419
+ "epoch": 0.07954162453656892,
420
+ "grad_norm": 0.4140625,
421
+ "learning_rate": 0.0007,
422
+ "loss": 2.484502410888672,
423
+ "step": 1180
424
+ },
425
+ {
426
+ "epoch": 0.08088978766430738,
427
+ "grad_norm": 0.51171875,
428
+ "learning_rate": 0.0007,
429
+ "loss": 2.476539421081543,
430
+ "step": 1200
431
+ },
432
+ {
433
+ "epoch": 0.08223795079204584,
434
+ "grad_norm": 0.41796875,
435
+ "learning_rate": 0.0007,
436
+ "loss": 2.463785171508789,
437
+ "step": 1220
438
+ },
439
+ {
440
+ "epoch": 0.0835861139197843,
441
+ "grad_norm": 0.390625,
442
+ "learning_rate": 0.0007,
443
+ "loss": 2.4509374618530275,
444
+ "step": 1240
445
+ },
446
+ {
447
+ "epoch": 0.08493427704752275,
448
+ "grad_norm": 0.4453125,
449
+ "learning_rate": 0.0007,
450
+ "loss": 2.4406501770019533,
451
+ "step": 1260
452
+ },
453
+ {
454
+ "epoch": 0.08628244017526121,
455
+ "grad_norm": 0.41796875,
456
+ "learning_rate": 0.0007,
457
+ "loss": 2.423234748840332,
458
+ "step": 1280
459
+ },
460
+ {
461
+ "epoch": 0.08763060330299967,
462
+ "grad_norm": 0.4453125,
463
+ "learning_rate": 0.0007,
464
+ "loss": 2.415129280090332,
465
+ "step": 1300
466
+ },
467
+ {
468
+ "epoch": 0.08897876643073811,
469
+ "grad_norm": 0.41015625,
470
+ "learning_rate": 0.0007,
471
+ "loss": 2.4075996398925783,
472
+ "step": 1320
473
+ },
474
+ {
475
+ "epoch": 0.09032692955847657,
476
+ "grad_norm": 0.412109375,
477
+ "learning_rate": 0.0007,
478
+ "loss": 2.396971321105957,
479
+ "step": 1340
480
+ },
481
+ {
482
+ "epoch": 0.09167509268621503,
483
+ "grad_norm": 0.431640625,
484
+ "learning_rate": 0.0007,
485
+ "loss": 2.398550605773926,
486
+ "step": 1360
487
+ },
488
+ {
489
+ "epoch": 0.09302325581395349,
490
+ "grad_norm": 0.40234375,
491
+ "learning_rate": 0.0007,
492
+ "loss": 2.3772783279418945,
493
+ "step": 1380
494
+ },
495
+ {
496
+ "epoch": 0.09437141894169195,
497
+ "grad_norm": 0.40625,
498
+ "learning_rate": 0.0007,
499
+ "loss": 2.3700124740600588,
500
+ "step": 1400
501
+ },
502
+ {
503
+ "epoch": 0.0957195820694304,
504
+ "grad_norm": 0.421875,
505
+ "learning_rate": 0.0007,
506
+ "loss": 2.3814950942993165,
507
+ "step": 1420
508
+ },
509
+ {
510
+ "epoch": 0.09706774519716886,
511
+ "grad_norm": 0.431640625,
512
+ "learning_rate": 0.0007,
513
+ "loss": 2.3590816497802733,
514
+ "step": 1440
515
+ },
516
+ {
517
+ "epoch": 0.09841590832490732,
518
+ "grad_norm": 0.470703125,
519
+ "learning_rate": 0.0007,
520
+ "loss": 2.347888374328613,
521
+ "step": 1460
522
+ },
523
+ {
524
+ "epoch": 0.09976407145264576,
525
+ "grad_norm": 0.400390625,
526
+ "learning_rate": 0.0007,
527
+ "loss": 2.3355857849121096,
528
+ "step": 1480
529
+ },
530
+ {
531
+ "epoch": 0.10111223458038422,
532
+ "grad_norm": 0.4140625,
533
+ "learning_rate": 0.0007,
534
+ "loss": 2.3234769821166994,
535
+ "step": 1500
536
+ },
537
+ {
538
+ "epoch": 0.10246039770812268,
539
+ "grad_norm": 0.4375,
540
+ "learning_rate": 0.0007,
541
+ "loss": 2.327121543884277,
542
+ "step": 1520
543
+ },
544
+ {
545
+ "epoch": 0.10380856083586114,
546
+ "grad_norm": 0.408203125,
547
+ "learning_rate": 0.0007,
548
+ "loss": 2.3147512435913087,
549
+ "step": 1540
550
+ },
551
+ {
552
+ "epoch": 0.1051567239635996,
553
+ "grad_norm": 0.408203125,
554
+ "learning_rate": 0.0007,
555
+ "loss": 2.3023752212524413,
556
+ "step": 1560
557
+ },
558
+ {
559
+ "epoch": 0.10650488709133805,
560
+ "grad_norm": 0.408203125,
561
+ "learning_rate": 0.0007,
562
+ "loss": 2.3237445831298826,
563
+ "step": 1580
564
+ },
565
+ {
566
+ "epoch": 0.10785305021907651,
567
+ "grad_norm": 0.400390625,
568
+ "learning_rate": 0.0007,
569
+ "loss": 2.288845252990723,
570
+ "step": 1600
571
+ },
572
+ {
573
+ "epoch": 0.10920121334681497,
574
+ "grad_norm": 0.4140625,
575
+ "learning_rate": 0.0007,
576
+ "loss": 2.28785400390625,
577
+ "step": 1620
578
+ },
579
+ {
580
+ "epoch": 0.11054937647455342,
581
+ "grad_norm": 0.416015625,
582
+ "learning_rate": 0.0007,
583
+ "loss": 2.300959587097168,
584
+ "step": 1640
585
+ },
586
+ {
587
+ "epoch": 0.11189753960229187,
588
+ "grad_norm": 0.380859375,
589
+ "learning_rate": 0.0007,
590
+ "loss": 2.284528350830078,
591
+ "step": 1660
592
+ },
593
+ {
594
+ "epoch": 0.11324570273003033,
595
+ "grad_norm": 0.408203125,
596
+ "learning_rate": 0.0007,
597
+ "loss": 2.267934799194336,
598
+ "step": 1680
599
+ },
600
+ {
601
+ "epoch": 0.11459386585776879,
602
+ "grad_norm": 0.423828125,
603
+ "learning_rate": 0.0007,
604
+ "loss": 2.2802879333496096,
605
+ "step": 1700
606
+ },
607
+ {
608
+ "epoch": 0.11594202898550725,
609
+ "grad_norm": 0.4375,
610
+ "learning_rate": 0.0007,
611
+ "loss": 2.258160400390625,
612
+ "step": 1720
613
+ },
614
+ {
615
+ "epoch": 0.1172901921132457,
616
+ "grad_norm": 0.39453125,
617
+ "learning_rate": 0.0007,
618
+ "loss": 2.246730995178223,
619
+ "step": 1740
620
+ },
621
+ {
622
+ "epoch": 0.11863835524098416,
623
+ "grad_norm": 0.419921875,
624
+ "learning_rate": 0.0007,
625
+ "loss": 2.252003479003906,
626
+ "step": 1760
627
+ },
628
+ {
629
+ "epoch": 0.11998651836872262,
630
+ "grad_norm": 0.38671875,
631
+ "learning_rate": 0.0007,
632
+ "loss": 2.265170669555664,
633
+ "step": 1780
634
+ },
635
+ {
636
+ "epoch": 0.12133468149646107,
637
+ "grad_norm": 0.408203125,
638
+ "learning_rate": 0.0007,
639
+ "loss": 2.2610910415649412,
640
+ "step": 1800
641
+ },
642
+ {
643
+ "epoch": 0.12268284462419952,
644
+ "grad_norm": 0.41015625,
645
+ "learning_rate": 0.0007,
646
+ "loss": 2.2539751052856447,
647
+ "step": 1820
648
+ },
649
+ {
650
+ "epoch": 0.12403100775193798,
651
+ "grad_norm": 0.404296875,
652
+ "learning_rate": 0.0007,
653
+ "loss": 2.220688819885254,
654
+ "step": 1840
655
+ },
656
+ {
657
+ "epoch": 0.12537917087967643,
658
+ "grad_norm": 0.376953125,
659
+ "learning_rate": 0.0007,
660
+ "loss": 2.222416305541992,
661
+ "step": 1860
662
+ },
663
+ {
664
+ "epoch": 0.1267273340074149,
665
+ "grad_norm": 0.3828125,
666
+ "learning_rate": 0.0007,
667
+ "loss": 2.2135461807250976,
668
+ "step": 1880
669
+ },
670
+ {
671
+ "epoch": 0.12807549713515334,
672
+ "grad_norm": 0.39453125,
673
+ "learning_rate": 0.0007,
674
+ "loss": 2.2209815979003906,
675
+ "step": 1900
676
+ },
677
+ {
678
+ "epoch": 0.12942366026289182,
679
+ "grad_norm": 0.39453125,
680
+ "learning_rate": 0.0007,
681
+ "loss": 2.199201965332031,
682
+ "step": 1920
683
+ },
684
+ {
685
+ "epoch": 0.13077182339063026,
686
+ "grad_norm": 0.41015625,
687
+ "learning_rate": 0.0007,
688
+ "loss": 2.2096010208129884,
689
+ "step": 1940
690
+ },
691
+ {
692
+ "epoch": 0.13211998651836873,
693
+ "grad_norm": 0.396484375,
694
+ "learning_rate": 0.0007,
695
+ "loss": 2.1849014282226564,
696
+ "step": 1960
697
+ },
698
+ {
699
+ "epoch": 0.13346814964610718,
700
+ "grad_norm": 0.421875,
701
+ "learning_rate": 0.0007,
702
+ "loss": 2.2073038101196287,
703
+ "step": 1980
704
+ },
705
+ {
706
+ "epoch": 0.13481631277384565,
707
+ "grad_norm": 0.373046875,
708
+ "learning_rate": 0.0007,
709
+ "loss": 2.1818737030029296,
710
+ "step": 2000
711
+ }
712
+ ],
713
+ "logging_steps": 20,
714
+ "max_steps": 2500,
715
+ "num_input_tokens_seen": 0,
716
+ "num_train_epochs": 1,
717
+ "save_steps": 100,
718
+ "stateful_callbacks": {
719
+ "TrainerControl": {
720
+ "args": {
721
+ "should_epoch_stop": false,
722
+ "should_evaluate": false,
723
+ "should_log": false,
724
+ "should_save": true,
725
+ "should_training_stop": false
726
+ },
727
+ "attributes": {}
728
+ }
729
+ },
730
+ "total_flos": 6452567605248000.0,
731
+ "train_batch_size": 64,
732
+ "trial_name": null,
733
+ "trial_params": null
734
+ }
zain/Activation/out/glu-waleed-100L_run/checkpoint-2000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c75eb6d8bf9c953837c430a69915c00b95cc26c05d95e0643e1a66470a65c0cf
3
+ size 4920
zain/Activation/out/glu-waleed-100L_run/checkpoint-2100/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "waleed",
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-waleed-100L_run/checkpoint-2100/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e74e24ea87bf26165f2b97600c7524b4d44d50480ffba6dabfe43a5ffe965782
3
+ size 33967272
zain/Activation/out/glu-waleed-100L_run/checkpoint-2100/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:659c035f2a3dc31571af28fcbb1d3d1ca351e3770477cb509f5964f07fa47856
3
+ size 68504996
zain/Activation/out/glu-waleed-100L_run/checkpoint-2100/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-waleed-100L_run/checkpoint-2100/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:985eadbe03d32df01f608a67ee99d09997f8b557ed94b6768e8445a622bb3975
3
+ size 1064
zain/Activation/out/glu-waleed-100L_run/checkpoint-2100/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-waleed-100L_run/checkpoint-2100/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-waleed-100L_run/checkpoint-2100/trainer_state.json ADDED
@@ -0,0 +1,769 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.14155712841253792,
6
+ "eval_steps": 2498,
7
+ "global_step": 2100,
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": 2.125,
15
+ "learning_rate": 2.66e-05,
16
+ "loss": 8.299404907226563,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.3125,
22
+ "learning_rate": 5.46e-05,
23
+ "loss": 8.036402130126953,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 1.2734375,
29
+ "learning_rate": 8.259999999999999e-05,
30
+ "loss": 7.834580993652343,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 1.265625,
36
+ "learning_rate": 0.0001106,
37
+ "loss": 7.5719047546386715,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 1.1953125,
43
+ "learning_rate": 0.0001386,
44
+ "loss": 7.237163543701172,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 1.1484375,
50
+ "learning_rate": 0.00016659999999999998,
51
+ "loss": 6.864958190917969,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 1.015625,
57
+ "learning_rate": 0.00019460000000000001,
58
+ "loss": 6.509477996826172,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 1.0234375,
64
+ "learning_rate": 0.0002226,
65
+ "loss": 6.204301452636718,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 1.0859375,
71
+ "learning_rate": 0.00025059999999999997,
72
+ "loss": 5.940593719482422,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 1.2109375,
78
+ "learning_rate": 0.0002786,
79
+ "loss": 5.676498794555664,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.01482979440512302,
84
+ "grad_norm": 0.8046875,
85
+ "learning_rate": 0.00030659999999999997,
86
+ "loss": 5.412863540649414,
87
+ "step": 220
88
+ },
89
+ {
90
+ "epoch": 0.016177957532861477,
91
+ "grad_norm": 0.71875,
92
+ "learning_rate": 0.0003346,
93
+ "loss": 5.1301216125488285,
94
+ "step": 240
95
+ },
96
+ {
97
+ "epoch": 0.01752612066059993,
98
+ "grad_norm": 1.046875,
99
+ "learning_rate": 0.00036260000000000003,
100
+ "loss": 4.88237190246582,
101
+ "step": 260
102
+ },
103
+ {
104
+ "epoch": 0.01887428378833839,
105
+ "grad_norm": 0.75,
106
+ "learning_rate": 0.0003906,
107
+ "loss": 4.643666076660156,
108
+ "step": 280
109
+ },
110
+ {
111
+ "epoch": 0.020222446916076844,
112
+ "grad_norm": 0.83984375,
113
+ "learning_rate": 0.0004186,
114
+ "loss": 4.441861343383789,
115
+ "step": 300
116
+ },
117
+ {
118
+ "epoch": 0.021570610043815303,
119
+ "grad_norm": 0.8125,
120
+ "learning_rate": 0.0004466,
121
+ "loss": 4.25420036315918,
122
+ "step": 320
123
+ },
124
+ {
125
+ "epoch": 0.022918773171553757,
126
+ "grad_norm": 0.62109375,
127
+ "learning_rate": 0.00047460000000000004,
128
+ "loss": 4.097484970092774,
129
+ "step": 340
130
+ },
131
+ {
132
+ "epoch": 0.024266936299292215,
133
+ "grad_norm": 0.6015625,
134
+ "learning_rate": 0.0005026,
135
+ "loss": 3.9764480590820312,
136
+ "step": 360
137
+ },
138
+ {
139
+ "epoch": 0.02561509942703067,
140
+ "grad_norm": 0.9140625,
141
+ "learning_rate": 0.0005306,
142
+ "loss": 3.854802703857422,
143
+ "step": 380
144
+ },
145
+ {
146
+ "epoch": 0.026963262554769128,
147
+ "grad_norm": 0.56640625,
148
+ "learning_rate": 0.0005586,
149
+ "loss": 3.7329875946044924,
150
+ "step": 400
151
+ },
152
+ {
153
+ "epoch": 0.028311425682507583,
154
+ "grad_norm": 0.53515625,
155
+ "learning_rate": 0.0005866,
156
+ "loss": 3.6484100341796877,
157
+ "step": 420
158
+ },
159
+ {
160
+ "epoch": 0.02965958881024604,
161
+ "grad_norm": 0.5390625,
162
+ "learning_rate": 0.0006146,
163
+ "loss": 3.595387268066406,
164
+ "step": 440
165
+ },
166
+ {
167
+ "epoch": 0.031007751937984496,
168
+ "grad_norm": 0.55859375,
169
+ "learning_rate": 0.0006426,
170
+ "loss": 3.498302459716797,
171
+ "step": 460
172
+ },
173
+ {
174
+ "epoch": 0.032355915065722954,
175
+ "grad_norm": 0.5859375,
176
+ "learning_rate": 0.0006705999999999999,
177
+ "loss": 3.4313674926757813,
178
+ "step": 480
179
+ },
180
+ {
181
+ "epoch": 0.03370407819346141,
182
+ "grad_norm": 0.61328125,
183
+ "learning_rate": 0.0006986,
184
+ "loss": 3.3613121032714846,
185
+ "step": 500
186
+ },
187
+ {
188
+ "epoch": 0.03505224132119986,
189
+ "grad_norm": 0.474609375,
190
+ "learning_rate": 0.0007,
191
+ "loss": 3.3039566040039063,
192
+ "step": 520
193
+ },
194
+ {
195
+ "epoch": 0.03640040444893832,
196
+ "grad_norm": 0.578125,
197
+ "learning_rate": 0.0007,
198
+ "loss": 3.2328567504882812,
199
+ "step": 540
200
+ },
201
+ {
202
+ "epoch": 0.03774856757667678,
203
+ "grad_norm": 0.64453125,
204
+ "learning_rate": 0.0007,
205
+ "loss": 3.187940788269043,
206
+ "step": 560
207
+ },
208
+ {
209
+ "epoch": 0.03909673070441524,
210
+ "grad_norm": 0.462890625,
211
+ "learning_rate": 0.0007,
212
+ "loss": 3.133651542663574,
213
+ "step": 580
214
+ },
215
+ {
216
+ "epoch": 0.04044489383215369,
217
+ "grad_norm": 0.5,
218
+ "learning_rate": 0.0007,
219
+ "loss": 3.1042255401611327,
220
+ "step": 600
221
+ },
222
+ {
223
+ "epoch": 0.04179305695989215,
224
+ "grad_norm": 0.578125,
225
+ "learning_rate": 0.0007,
226
+ "loss": 3.04781551361084,
227
+ "step": 620
228
+ },
229
+ {
230
+ "epoch": 0.043141220087630605,
231
+ "grad_norm": 0.453125,
232
+ "learning_rate": 0.0007,
233
+ "loss": 3.0298320770263674,
234
+ "step": 640
235
+ },
236
+ {
237
+ "epoch": 0.044489383215369056,
238
+ "grad_norm": 0.46875,
239
+ "learning_rate": 0.0007,
240
+ "loss": 2.9790300369262694,
241
+ "step": 660
242
+ },
243
+ {
244
+ "epoch": 0.045837546343107514,
245
+ "grad_norm": 0.46875,
246
+ "learning_rate": 0.0007,
247
+ "loss": 2.9566295623779295,
248
+ "step": 680
249
+ },
250
+ {
251
+ "epoch": 0.04718570947084597,
252
+ "grad_norm": 0.439453125,
253
+ "learning_rate": 0.0007,
254
+ "loss": 2.916979217529297,
255
+ "step": 700
256
+ },
257
+ {
258
+ "epoch": 0.04853387259858443,
259
+ "grad_norm": 0.41796875,
260
+ "learning_rate": 0.0007,
261
+ "loss": 2.8888761520385744,
262
+ "step": 720
263
+ },
264
+ {
265
+ "epoch": 0.04988203572632288,
266
+ "grad_norm": 0.4765625,
267
+ "learning_rate": 0.0007,
268
+ "loss": 2.855138969421387,
269
+ "step": 740
270
+ },
271
+ {
272
+ "epoch": 0.05123019885406134,
273
+ "grad_norm": 0.46875,
274
+ "learning_rate": 0.0007,
275
+ "loss": 2.841838836669922,
276
+ "step": 760
277
+ },
278
+ {
279
+ "epoch": 0.0525783619817998,
280
+ "grad_norm": 0.43359375,
281
+ "learning_rate": 0.0007,
282
+ "loss": 2.8384235382080076,
283
+ "step": 780
284
+ },
285
+ {
286
+ "epoch": 0.053926525109538256,
287
+ "grad_norm": 0.447265625,
288
+ "learning_rate": 0.0007,
289
+ "loss": 2.7947757720947264,
290
+ "step": 800
291
+ },
292
+ {
293
+ "epoch": 0.05527468823727671,
294
+ "grad_norm": 0.478515625,
295
+ "learning_rate": 0.0007,
296
+ "loss": 2.7847801208496095,
297
+ "step": 820
298
+ },
299
+ {
300
+ "epoch": 0.056622851365015166,
301
+ "grad_norm": 0.44140625,
302
+ "learning_rate": 0.0007,
303
+ "loss": 2.7601039886474608,
304
+ "step": 840
305
+ },
306
+ {
307
+ "epoch": 0.057971014492753624,
308
+ "grad_norm": 0.484375,
309
+ "learning_rate": 0.0007,
310
+ "loss": 2.716906929016113,
311
+ "step": 860
312
+ },
313
+ {
314
+ "epoch": 0.05931917762049208,
315
+ "grad_norm": 0.439453125,
316
+ "learning_rate": 0.0007,
317
+ "loss": 2.6968852996826174,
318
+ "step": 880
319
+ },
320
+ {
321
+ "epoch": 0.06066734074823053,
322
+ "grad_norm": 0.447265625,
323
+ "learning_rate": 0.0007,
324
+ "loss": 2.6847219467163086,
325
+ "step": 900
326
+ },
327
+ {
328
+ "epoch": 0.06201550387596899,
329
+ "grad_norm": 0.435546875,
330
+ "learning_rate": 0.0007,
331
+ "loss": 2.6599369049072266,
332
+ "step": 920
333
+ },
334
+ {
335
+ "epoch": 0.06336366700370745,
336
+ "grad_norm": 0.455078125,
337
+ "learning_rate": 0.0007,
338
+ "loss": 2.6465431213378907,
339
+ "step": 940
340
+ },
341
+ {
342
+ "epoch": 0.06471183013144591,
343
+ "grad_norm": 0.455078125,
344
+ "learning_rate": 0.0007,
345
+ "loss": 2.623439979553223,
346
+ "step": 960
347
+ },
348
+ {
349
+ "epoch": 0.06605999325918437,
350
+ "grad_norm": 0.462890625,
351
+ "learning_rate": 0.0007,
352
+ "loss": 2.609505271911621,
353
+ "step": 980
354
+ },
355
+ {
356
+ "epoch": 0.06740815638692282,
357
+ "grad_norm": 0.48828125,
358
+ "learning_rate": 0.0007,
359
+ "loss": 2.610812187194824,
360
+ "step": 1000
361
+ },
362
+ {
363
+ "epoch": 0.06875631951466127,
364
+ "grad_norm": 0.423828125,
365
+ "learning_rate": 0.0007,
366
+ "loss": 2.581887054443359,
367
+ "step": 1020
368
+ },
369
+ {
370
+ "epoch": 0.07010448264239973,
371
+ "grad_norm": 0.4453125,
372
+ "learning_rate": 0.0007,
373
+ "loss": 2.5679080963134764,
374
+ "step": 1040
375
+ },
376
+ {
377
+ "epoch": 0.07145264577013818,
378
+ "grad_norm": 0.447265625,
379
+ "learning_rate": 0.0007,
380
+ "loss": 2.55521240234375,
381
+ "step": 1060
382
+ },
383
+ {
384
+ "epoch": 0.07280080889787664,
385
+ "grad_norm": 0.4453125,
386
+ "learning_rate": 0.0007,
387
+ "loss": 2.536572074890137,
388
+ "step": 1080
389
+ },
390
+ {
391
+ "epoch": 0.0741489720256151,
392
+ "grad_norm": 0.431640625,
393
+ "learning_rate": 0.0007,
394
+ "loss": 2.5341920852661133,
395
+ "step": 1100
396
+ },
397
+ {
398
+ "epoch": 0.07549713515335356,
399
+ "grad_norm": 0.427734375,
400
+ "learning_rate": 0.0007,
401
+ "loss": 2.4994813919067385,
402
+ "step": 1120
403
+ },
404
+ {
405
+ "epoch": 0.07684529828109202,
406
+ "grad_norm": 0.421875,
407
+ "learning_rate": 0.0007,
408
+ "loss": 2.4886734008789064,
409
+ "step": 1140
410
+ },
411
+ {
412
+ "epoch": 0.07819346140883048,
413
+ "grad_norm": 0.419921875,
414
+ "learning_rate": 0.0007,
415
+ "loss": 2.4889175415039064,
416
+ "step": 1160
417
+ },
418
+ {
419
+ "epoch": 0.07954162453656892,
420
+ "grad_norm": 0.4140625,
421
+ "learning_rate": 0.0007,
422
+ "loss": 2.484502410888672,
423
+ "step": 1180
424
+ },
425
+ {
426
+ "epoch": 0.08088978766430738,
427
+ "grad_norm": 0.51171875,
428
+ "learning_rate": 0.0007,
429
+ "loss": 2.476539421081543,
430
+ "step": 1200
431
+ },
432
+ {
433
+ "epoch": 0.08223795079204584,
434
+ "grad_norm": 0.41796875,
435
+ "learning_rate": 0.0007,
436
+ "loss": 2.463785171508789,
437
+ "step": 1220
438
+ },
439
+ {
440
+ "epoch": 0.0835861139197843,
441
+ "grad_norm": 0.390625,
442
+ "learning_rate": 0.0007,
443
+ "loss": 2.4509374618530275,
444
+ "step": 1240
445
+ },
446
+ {
447
+ "epoch": 0.08493427704752275,
448
+ "grad_norm": 0.4453125,
449
+ "learning_rate": 0.0007,
450
+ "loss": 2.4406501770019533,
451
+ "step": 1260
452
+ },
453
+ {
454
+ "epoch": 0.08628244017526121,
455
+ "grad_norm": 0.41796875,
456
+ "learning_rate": 0.0007,
457
+ "loss": 2.423234748840332,
458
+ "step": 1280
459
+ },
460
+ {
461
+ "epoch": 0.08763060330299967,
462
+ "grad_norm": 0.4453125,
463
+ "learning_rate": 0.0007,
464
+ "loss": 2.415129280090332,
465
+ "step": 1300
466
+ },
467
+ {
468
+ "epoch": 0.08897876643073811,
469
+ "grad_norm": 0.41015625,
470
+ "learning_rate": 0.0007,
471
+ "loss": 2.4075996398925783,
472
+ "step": 1320
473
+ },
474
+ {
475
+ "epoch": 0.09032692955847657,
476
+ "grad_norm": 0.412109375,
477
+ "learning_rate": 0.0007,
478
+ "loss": 2.396971321105957,
479
+ "step": 1340
480
+ },
481
+ {
482
+ "epoch": 0.09167509268621503,
483
+ "grad_norm": 0.431640625,
484
+ "learning_rate": 0.0007,
485
+ "loss": 2.398550605773926,
486
+ "step": 1360
487
+ },
488
+ {
489
+ "epoch": 0.09302325581395349,
490
+ "grad_norm": 0.40234375,
491
+ "learning_rate": 0.0007,
492
+ "loss": 2.3772783279418945,
493
+ "step": 1380
494
+ },
495
+ {
496
+ "epoch": 0.09437141894169195,
497
+ "grad_norm": 0.40625,
498
+ "learning_rate": 0.0007,
499
+ "loss": 2.3700124740600588,
500
+ "step": 1400
501
+ },
502
+ {
503
+ "epoch": 0.0957195820694304,
504
+ "grad_norm": 0.421875,
505
+ "learning_rate": 0.0007,
506
+ "loss": 2.3814950942993165,
507
+ "step": 1420
508
+ },
509
+ {
510
+ "epoch": 0.09706774519716886,
511
+ "grad_norm": 0.431640625,
512
+ "learning_rate": 0.0007,
513
+ "loss": 2.3590816497802733,
514
+ "step": 1440
515
+ },
516
+ {
517
+ "epoch": 0.09841590832490732,
518
+ "grad_norm": 0.470703125,
519
+ "learning_rate": 0.0007,
520
+ "loss": 2.347888374328613,
521
+ "step": 1460
522
+ },
523
+ {
524
+ "epoch": 0.09976407145264576,
525
+ "grad_norm": 0.400390625,
526
+ "learning_rate": 0.0007,
527
+ "loss": 2.3355857849121096,
528
+ "step": 1480
529
+ },
530
+ {
531
+ "epoch": 0.10111223458038422,
532
+ "grad_norm": 0.4140625,
533
+ "learning_rate": 0.0007,
534
+ "loss": 2.3234769821166994,
535
+ "step": 1500
536
+ },
537
+ {
538
+ "epoch": 0.10246039770812268,
539
+ "grad_norm": 0.4375,
540
+ "learning_rate": 0.0007,
541
+ "loss": 2.327121543884277,
542
+ "step": 1520
543
+ },
544
+ {
545
+ "epoch": 0.10380856083586114,
546
+ "grad_norm": 0.408203125,
547
+ "learning_rate": 0.0007,
548
+ "loss": 2.3147512435913087,
549
+ "step": 1540
550
+ },
551
+ {
552
+ "epoch": 0.1051567239635996,
553
+ "grad_norm": 0.408203125,
554
+ "learning_rate": 0.0007,
555
+ "loss": 2.3023752212524413,
556
+ "step": 1560
557
+ },
558
+ {
559
+ "epoch": 0.10650488709133805,
560
+ "grad_norm": 0.408203125,
561
+ "learning_rate": 0.0007,
562
+ "loss": 2.3237445831298826,
563
+ "step": 1580
564
+ },
565
+ {
566
+ "epoch": 0.10785305021907651,
567
+ "grad_norm": 0.400390625,
568
+ "learning_rate": 0.0007,
569
+ "loss": 2.288845252990723,
570
+ "step": 1600
571
+ },
572
+ {
573
+ "epoch": 0.10920121334681497,
574
+ "grad_norm": 0.4140625,
575
+ "learning_rate": 0.0007,
576
+ "loss": 2.28785400390625,
577
+ "step": 1620
578
+ },
579
+ {
580
+ "epoch": 0.11054937647455342,
581
+ "grad_norm": 0.416015625,
582
+ "learning_rate": 0.0007,
583
+ "loss": 2.300959587097168,
584
+ "step": 1640
585
+ },
586
+ {
587
+ "epoch": 0.11189753960229187,
588
+ "grad_norm": 0.380859375,
589
+ "learning_rate": 0.0007,
590
+ "loss": 2.284528350830078,
591
+ "step": 1660
592
+ },
593
+ {
594
+ "epoch": 0.11324570273003033,
595
+ "grad_norm": 0.408203125,
596
+ "learning_rate": 0.0007,
597
+ "loss": 2.267934799194336,
598
+ "step": 1680
599
+ },
600
+ {
601
+ "epoch": 0.11459386585776879,
602
+ "grad_norm": 0.423828125,
603
+ "learning_rate": 0.0007,
604
+ "loss": 2.2802879333496096,
605
+ "step": 1700
606
+ },
607
+ {
608
+ "epoch": 0.11594202898550725,
609
+ "grad_norm": 0.4375,
610
+ "learning_rate": 0.0007,
611
+ "loss": 2.258160400390625,
612
+ "step": 1720
613
+ },
614
+ {
615
+ "epoch": 0.1172901921132457,
616
+ "grad_norm": 0.39453125,
617
+ "learning_rate": 0.0007,
618
+ "loss": 2.246730995178223,
619
+ "step": 1740
620
+ },
621
+ {
622
+ "epoch": 0.11863835524098416,
623
+ "grad_norm": 0.419921875,
624
+ "learning_rate": 0.0007,
625
+ "loss": 2.252003479003906,
626
+ "step": 1760
627
+ },
628
+ {
629
+ "epoch": 0.11998651836872262,
630
+ "grad_norm": 0.38671875,
631
+ "learning_rate": 0.0007,
632
+ "loss": 2.265170669555664,
633
+ "step": 1780
634
+ },
635
+ {
636
+ "epoch": 0.12133468149646107,
637
+ "grad_norm": 0.408203125,
638
+ "learning_rate": 0.0007,
639
+ "loss": 2.2610910415649412,
640
+ "step": 1800
641
+ },
642
+ {
643
+ "epoch": 0.12268284462419952,
644
+ "grad_norm": 0.41015625,
645
+ "learning_rate": 0.0007,
646
+ "loss": 2.2539751052856447,
647
+ "step": 1820
648
+ },
649
+ {
650
+ "epoch": 0.12403100775193798,
651
+ "grad_norm": 0.404296875,
652
+ "learning_rate": 0.0007,
653
+ "loss": 2.220688819885254,
654
+ "step": 1840
655
+ },
656
+ {
657
+ "epoch": 0.12537917087967643,
658
+ "grad_norm": 0.376953125,
659
+ "learning_rate": 0.0007,
660
+ "loss": 2.222416305541992,
661
+ "step": 1860
662
+ },
663
+ {
664
+ "epoch": 0.1267273340074149,
665
+ "grad_norm": 0.3828125,
666
+ "learning_rate": 0.0007,
667
+ "loss": 2.2135461807250976,
668
+ "step": 1880
669
+ },
670
+ {
671
+ "epoch": 0.12807549713515334,
672
+ "grad_norm": 0.39453125,
673
+ "learning_rate": 0.0007,
674
+ "loss": 2.2209815979003906,
675
+ "step": 1900
676
+ },
677
+ {
678
+ "epoch": 0.12942366026289182,
679
+ "grad_norm": 0.39453125,
680
+ "learning_rate": 0.0007,
681
+ "loss": 2.199201965332031,
682
+ "step": 1920
683
+ },
684
+ {
685
+ "epoch": 0.13077182339063026,
686
+ "grad_norm": 0.41015625,
687
+ "learning_rate": 0.0007,
688
+ "loss": 2.2096010208129884,
689
+ "step": 1940
690
+ },
691
+ {
692
+ "epoch": 0.13211998651836873,
693
+ "grad_norm": 0.396484375,
694
+ "learning_rate": 0.0007,
695
+ "loss": 2.1849014282226564,
696
+ "step": 1960
697
+ },
698
+ {
699
+ "epoch": 0.13346814964610718,
700
+ "grad_norm": 0.421875,
701
+ "learning_rate": 0.0007,
702
+ "loss": 2.2073038101196287,
703
+ "step": 1980
704
+ },
705
+ {
706
+ "epoch": 0.13481631277384565,
707
+ "grad_norm": 0.373046875,
708
+ "learning_rate": 0.0007,
709
+ "loss": 2.1818737030029296,
710
+ "step": 2000
711
+ },
712
+ {
713
+ "epoch": 0.1361644759015841,
714
+ "grad_norm": 0.384765625,
715
+ "learning_rate": 0.0007,
716
+ "loss": 2.1887689590454102,
717
+ "step": 2020
718
+ },
719
+ {
720
+ "epoch": 0.13751263902932254,
721
+ "grad_norm": 0.4296875,
722
+ "learning_rate": 0.0007,
723
+ "loss": 2.1714288711547853,
724
+ "step": 2040
725
+ },
726
+ {
727
+ "epoch": 0.138860802157061,
728
+ "grad_norm": 0.392578125,
729
+ "learning_rate": 0.0007,
730
+ "loss": 2.1905704498291017,
731
+ "step": 2060
732
+ },
733
+ {
734
+ "epoch": 0.14020896528479945,
735
+ "grad_norm": 0.388671875,
736
+ "learning_rate": 0.0007,
737
+ "loss": 2.168014717102051,
738
+ "step": 2080
739
+ },
740
+ {
741
+ "epoch": 0.14155712841253792,
742
+ "grad_norm": 0.37890625,
743
+ "learning_rate": 0.0007,
744
+ "loss": 2.157928466796875,
745
+ "step": 2100
746
+ }
747
+ ],
748
+ "logging_steps": 20,
749
+ "max_steps": 2500,
750
+ "num_input_tokens_seen": 0,
751
+ "num_train_epochs": 1,
752
+ "save_steps": 100,
753
+ "stateful_callbacks": {
754
+ "TrainerControl": {
755
+ "args": {
756
+ "should_epoch_stop": false,
757
+ "should_evaluate": false,
758
+ "should_log": false,
759
+ "should_save": true,
760
+ "should_training_stop": false
761
+ },
762
+ "attributes": {}
763
+ }
764
+ },
765
+ "total_flos": 6775195985510400.0,
766
+ "train_batch_size": 64,
767
+ "trial_name": null,
768
+ "trial_params": null
769
+ }
zain/Activation/out/glu-waleed-100L_run/checkpoint-2100/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c75eb6d8bf9c953837c430a69915c00b95cc26c05d95e0643e1a66470a65c0cf
3
+ size 4920
zain/Activation/out/glu-waleed-100L_run/checkpoint-2200/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "waleed",
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-waleed-100L_run/checkpoint-2200/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b8399f69af5b80747be3025139c03644862231344793b2f63d9a30d6a389dec2
3
+ size 33967272
zain/Activation/out/glu-waleed-100L_run/checkpoint-2200/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:496a1122430b423455ffe69cc685f286ed0c7b0f9558885f04815bbf9672fa13
3
+ size 68504996
zain/Activation/out/glu-waleed-100L_run/checkpoint-2200/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-waleed-100L_run/checkpoint-2200/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bb38d8fe479c6ce36db4ed357fb0375cafd5d59c32c20acad02585958763071c
3
+ size 1064
zain/Activation/out/glu-waleed-100L_run/checkpoint-2200/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-waleed-100L_run/checkpoint-2200/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-waleed-100L_run/checkpoint-2200/trainer_state.json ADDED
@@ -0,0 +1,804 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.1482979440512302,
6
+ "eval_steps": 2498,
7
+ "global_step": 2200,
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": 2.125,
15
+ "learning_rate": 2.66e-05,
16
+ "loss": 8.299404907226563,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.3125,
22
+ "learning_rate": 5.46e-05,
23
+ "loss": 8.036402130126953,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 1.2734375,
29
+ "learning_rate": 8.259999999999999e-05,
30
+ "loss": 7.834580993652343,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 1.265625,
36
+ "learning_rate": 0.0001106,
37
+ "loss": 7.5719047546386715,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 1.1953125,
43
+ "learning_rate": 0.0001386,
44
+ "loss": 7.237163543701172,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 1.1484375,
50
+ "learning_rate": 0.00016659999999999998,
51
+ "loss": 6.864958190917969,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 1.015625,
57
+ "learning_rate": 0.00019460000000000001,
58
+ "loss": 6.509477996826172,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 1.0234375,
64
+ "learning_rate": 0.0002226,
65
+ "loss": 6.204301452636718,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 1.0859375,
71
+ "learning_rate": 0.00025059999999999997,
72
+ "loss": 5.940593719482422,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 1.2109375,
78
+ "learning_rate": 0.0002786,
79
+ "loss": 5.676498794555664,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.01482979440512302,
84
+ "grad_norm": 0.8046875,
85
+ "learning_rate": 0.00030659999999999997,
86
+ "loss": 5.412863540649414,
87
+ "step": 220
88
+ },
89
+ {
90
+ "epoch": 0.016177957532861477,
91
+ "grad_norm": 0.71875,
92
+ "learning_rate": 0.0003346,
93
+ "loss": 5.1301216125488285,
94
+ "step": 240
95
+ },
96
+ {
97
+ "epoch": 0.01752612066059993,
98
+ "grad_norm": 1.046875,
99
+ "learning_rate": 0.00036260000000000003,
100
+ "loss": 4.88237190246582,
101
+ "step": 260
102
+ },
103
+ {
104
+ "epoch": 0.01887428378833839,
105
+ "grad_norm": 0.75,
106
+ "learning_rate": 0.0003906,
107
+ "loss": 4.643666076660156,
108
+ "step": 280
109
+ },
110
+ {
111
+ "epoch": 0.020222446916076844,
112
+ "grad_norm": 0.83984375,
113
+ "learning_rate": 0.0004186,
114
+ "loss": 4.441861343383789,
115
+ "step": 300
116
+ },
117
+ {
118
+ "epoch": 0.021570610043815303,
119
+ "grad_norm": 0.8125,
120
+ "learning_rate": 0.0004466,
121
+ "loss": 4.25420036315918,
122
+ "step": 320
123
+ },
124
+ {
125
+ "epoch": 0.022918773171553757,
126
+ "grad_norm": 0.62109375,
127
+ "learning_rate": 0.00047460000000000004,
128
+ "loss": 4.097484970092774,
129
+ "step": 340
130
+ },
131
+ {
132
+ "epoch": 0.024266936299292215,
133
+ "grad_norm": 0.6015625,
134
+ "learning_rate": 0.0005026,
135
+ "loss": 3.9764480590820312,
136
+ "step": 360
137
+ },
138
+ {
139
+ "epoch": 0.02561509942703067,
140
+ "grad_norm": 0.9140625,
141
+ "learning_rate": 0.0005306,
142
+ "loss": 3.854802703857422,
143
+ "step": 380
144
+ },
145
+ {
146
+ "epoch": 0.026963262554769128,
147
+ "grad_norm": 0.56640625,
148
+ "learning_rate": 0.0005586,
149
+ "loss": 3.7329875946044924,
150
+ "step": 400
151
+ },
152
+ {
153
+ "epoch": 0.028311425682507583,
154
+ "grad_norm": 0.53515625,
155
+ "learning_rate": 0.0005866,
156
+ "loss": 3.6484100341796877,
157
+ "step": 420
158
+ },
159
+ {
160
+ "epoch": 0.02965958881024604,
161
+ "grad_norm": 0.5390625,
162
+ "learning_rate": 0.0006146,
163
+ "loss": 3.595387268066406,
164
+ "step": 440
165
+ },
166
+ {
167
+ "epoch": 0.031007751937984496,
168
+ "grad_norm": 0.55859375,
169
+ "learning_rate": 0.0006426,
170
+ "loss": 3.498302459716797,
171
+ "step": 460
172
+ },
173
+ {
174
+ "epoch": 0.032355915065722954,
175
+ "grad_norm": 0.5859375,
176
+ "learning_rate": 0.0006705999999999999,
177
+ "loss": 3.4313674926757813,
178
+ "step": 480
179
+ },
180
+ {
181
+ "epoch": 0.03370407819346141,
182
+ "grad_norm": 0.61328125,
183
+ "learning_rate": 0.0006986,
184
+ "loss": 3.3613121032714846,
185
+ "step": 500
186
+ },
187
+ {
188
+ "epoch": 0.03505224132119986,
189
+ "grad_norm": 0.474609375,
190
+ "learning_rate": 0.0007,
191
+ "loss": 3.3039566040039063,
192
+ "step": 520
193
+ },
194
+ {
195
+ "epoch": 0.03640040444893832,
196
+ "grad_norm": 0.578125,
197
+ "learning_rate": 0.0007,
198
+ "loss": 3.2328567504882812,
199
+ "step": 540
200
+ },
201
+ {
202
+ "epoch": 0.03774856757667678,
203
+ "grad_norm": 0.64453125,
204
+ "learning_rate": 0.0007,
205
+ "loss": 3.187940788269043,
206
+ "step": 560
207
+ },
208
+ {
209
+ "epoch": 0.03909673070441524,
210
+ "grad_norm": 0.462890625,
211
+ "learning_rate": 0.0007,
212
+ "loss": 3.133651542663574,
213
+ "step": 580
214
+ },
215
+ {
216
+ "epoch": 0.04044489383215369,
217
+ "grad_norm": 0.5,
218
+ "learning_rate": 0.0007,
219
+ "loss": 3.1042255401611327,
220
+ "step": 600
221
+ },
222
+ {
223
+ "epoch": 0.04179305695989215,
224
+ "grad_norm": 0.578125,
225
+ "learning_rate": 0.0007,
226
+ "loss": 3.04781551361084,
227
+ "step": 620
228
+ },
229
+ {
230
+ "epoch": 0.043141220087630605,
231
+ "grad_norm": 0.453125,
232
+ "learning_rate": 0.0007,
233
+ "loss": 3.0298320770263674,
234
+ "step": 640
235
+ },
236
+ {
237
+ "epoch": 0.044489383215369056,
238
+ "grad_norm": 0.46875,
239
+ "learning_rate": 0.0007,
240
+ "loss": 2.9790300369262694,
241
+ "step": 660
242
+ },
243
+ {
244
+ "epoch": 0.045837546343107514,
245
+ "grad_norm": 0.46875,
246
+ "learning_rate": 0.0007,
247
+ "loss": 2.9566295623779295,
248
+ "step": 680
249
+ },
250
+ {
251
+ "epoch": 0.04718570947084597,
252
+ "grad_norm": 0.439453125,
253
+ "learning_rate": 0.0007,
254
+ "loss": 2.916979217529297,
255
+ "step": 700
256
+ },
257
+ {
258
+ "epoch": 0.04853387259858443,
259
+ "grad_norm": 0.41796875,
260
+ "learning_rate": 0.0007,
261
+ "loss": 2.8888761520385744,
262
+ "step": 720
263
+ },
264
+ {
265
+ "epoch": 0.04988203572632288,
266
+ "grad_norm": 0.4765625,
267
+ "learning_rate": 0.0007,
268
+ "loss": 2.855138969421387,
269
+ "step": 740
270
+ },
271
+ {
272
+ "epoch": 0.05123019885406134,
273
+ "grad_norm": 0.46875,
274
+ "learning_rate": 0.0007,
275
+ "loss": 2.841838836669922,
276
+ "step": 760
277
+ },
278
+ {
279
+ "epoch": 0.0525783619817998,
280
+ "grad_norm": 0.43359375,
281
+ "learning_rate": 0.0007,
282
+ "loss": 2.8384235382080076,
283
+ "step": 780
284
+ },
285
+ {
286
+ "epoch": 0.053926525109538256,
287
+ "grad_norm": 0.447265625,
288
+ "learning_rate": 0.0007,
289
+ "loss": 2.7947757720947264,
290
+ "step": 800
291
+ },
292
+ {
293
+ "epoch": 0.05527468823727671,
294
+ "grad_norm": 0.478515625,
295
+ "learning_rate": 0.0007,
296
+ "loss": 2.7847801208496095,
297
+ "step": 820
298
+ },
299
+ {
300
+ "epoch": 0.056622851365015166,
301
+ "grad_norm": 0.44140625,
302
+ "learning_rate": 0.0007,
303
+ "loss": 2.7601039886474608,
304
+ "step": 840
305
+ },
306
+ {
307
+ "epoch": 0.057971014492753624,
308
+ "grad_norm": 0.484375,
309
+ "learning_rate": 0.0007,
310
+ "loss": 2.716906929016113,
311
+ "step": 860
312
+ },
313
+ {
314
+ "epoch": 0.05931917762049208,
315
+ "grad_norm": 0.439453125,
316
+ "learning_rate": 0.0007,
317
+ "loss": 2.6968852996826174,
318
+ "step": 880
319
+ },
320
+ {
321
+ "epoch": 0.06066734074823053,
322
+ "grad_norm": 0.447265625,
323
+ "learning_rate": 0.0007,
324
+ "loss": 2.6847219467163086,
325
+ "step": 900
326
+ },
327
+ {
328
+ "epoch": 0.06201550387596899,
329
+ "grad_norm": 0.435546875,
330
+ "learning_rate": 0.0007,
331
+ "loss": 2.6599369049072266,
332
+ "step": 920
333
+ },
334
+ {
335
+ "epoch": 0.06336366700370745,
336
+ "grad_norm": 0.455078125,
337
+ "learning_rate": 0.0007,
338
+ "loss": 2.6465431213378907,
339
+ "step": 940
340
+ },
341
+ {
342
+ "epoch": 0.06471183013144591,
343
+ "grad_norm": 0.455078125,
344
+ "learning_rate": 0.0007,
345
+ "loss": 2.623439979553223,
346
+ "step": 960
347
+ },
348
+ {
349
+ "epoch": 0.06605999325918437,
350
+ "grad_norm": 0.462890625,
351
+ "learning_rate": 0.0007,
352
+ "loss": 2.609505271911621,
353
+ "step": 980
354
+ },
355
+ {
356
+ "epoch": 0.06740815638692282,
357
+ "grad_norm": 0.48828125,
358
+ "learning_rate": 0.0007,
359
+ "loss": 2.610812187194824,
360
+ "step": 1000
361
+ },
362
+ {
363
+ "epoch": 0.06875631951466127,
364
+ "grad_norm": 0.423828125,
365
+ "learning_rate": 0.0007,
366
+ "loss": 2.581887054443359,
367
+ "step": 1020
368
+ },
369
+ {
370
+ "epoch": 0.07010448264239973,
371
+ "grad_norm": 0.4453125,
372
+ "learning_rate": 0.0007,
373
+ "loss": 2.5679080963134764,
374
+ "step": 1040
375
+ },
376
+ {
377
+ "epoch": 0.07145264577013818,
378
+ "grad_norm": 0.447265625,
379
+ "learning_rate": 0.0007,
380
+ "loss": 2.55521240234375,
381
+ "step": 1060
382
+ },
383
+ {
384
+ "epoch": 0.07280080889787664,
385
+ "grad_norm": 0.4453125,
386
+ "learning_rate": 0.0007,
387
+ "loss": 2.536572074890137,
388
+ "step": 1080
389
+ },
390
+ {
391
+ "epoch": 0.0741489720256151,
392
+ "grad_norm": 0.431640625,
393
+ "learning_rate": 0.0007,
394
+ "loss": 2.5341920852661133,
395
+ "step": 1100
396
+ },
397
+ {
398
+ "epoch": 0.07549713515335356,
399
+ "grad_norm": 0.427734375,
400
+ "learning_rate": 0.0007,
401
+ "loss": 2.4994813919067385,
402
+ "step": 1120
403
+ },
404
+ {
405
+ "epoch": 0.07684529828109202,
406
+ "grad_norm": 0.421875,
407
+ "learning_rate": 0.0007,
408
+ "loss": 2.4886734008789064,
409
+ "step": 1140
410
+ },
411
+ {
412
+ "epoch": 0.07819346140883048,
413
+ "grad_norm": 0.419921875,
414
+ "learning_rate": 0.0007,
415
+ "loss": 2.4889175415039064,
416
+ "step": 1160
417
+ },
418
+ {
419
+ "epoch": 0.07954162453656892,
420
+ "grad_norm": 0.4140625,
421
+ "learning_rate": 0.0007,
422
+ "loss": 2.484502410888672,
423
+ "step": 1180
424
+ },
425
+ {
426
+ "epoch": 0.08088978766430738,
427
+ "grad_norm": 0.51171875,
428
+ "learning_rate": 0.0007,
429
+ "loss": 2.476539421081543,
430
+ "step": 1200
431
+ },
432
+ {
433
+ "epoch": 0.08223795079204584,
434
+ "grad_norm": 0.41796875,
435
+ "learning_rate": 0.0007,
436
+ "loss": 2.463785171508789,
437
+ "step": 1220
438
+ },
439
+ {
440
+ "epoch": 0.0835861139197843,
441
+ "grad_norm": 0.390625,
442
+ "learning_rate": 0.0007,
443
+ "loss": 2.4509374618530275,
444
+ "step": 1240
445
+ },
446
+ {
447
+ "epoch": 0.08493427704752275,
448
+ "grad_norm": 0.4453125,
449
+ "learning_rate": 0.0007,
450
+ "loss": 2.4406501770019533,
451
+ "step": 1260
452
+ },
453
+ {
454
+ "epoch": 0.08628244017526121,
455
+ "grad_norm": 0.41796875,
456
+ "learning_rate": 0.0007,
457
+ "loss": 2.423234748840332,
458
+ "step": 1280
459
+ },
460
+ {
461
+ "epoch": 0.08763060330299967,
462
+ "grad_norm": 0.4453125,
463
+ "learning_rate": 0.0007,
464
+ "loss": 2.415129280090332,
465
+ "step": 1300
466
+ },
467
+ {
468
+ "epoch": 0.08897876643073811,
469
+ "grad_norm": 0.41015625,
470
+ "learning_rate": 0.0007,
471
+ "loss": 2.4075996398925783,
472
+ "step": 1320
473
+ },
474
+ {
475
+ "epoch": 0.09032692955847657,
476
+ "grad_norm": 0.412109375,
477
+ "learning_rate": 0.0007,
478
+ "loss": 2.396971321105957,
479
+ "step": 1340
480
+ },
481
+ {
482
+ "epoch": 0.09167509268621503,
483
+ "grad_norm": 0.431640625,
484
+ "learning_rate": 0.0007,
485
+ "loss": 2.398550605773926,
486
+ "step": 1360
487
+ },
488
+ {
489
+ "epoch": 0.09302325581395349,
490
+ "grad_norm": 0.40234375,
491
+ "learning_rate": 0.0007,
492
+ "loss": 2.3772783279418945,
493
+ "step": 1380
494
+ },
495
+ {
496
+ "epoch": 0.09437141894169195,
497
+ "grad_norm": 0.40625,
498
+ "learning_rate": 0.0007,
499
+ "loss": 2.3700124740600588,
500
+ "step": 1400
501
+ },
502
+ {
503
+ "epoch": 0.0957195820694304,
504
+ "grad_norm": 0.421875,
505
+ "learning_rate": 0.0007,
506
+ "loss": 2.3814950942993165,
507
+ "step": 1420
508
+ },
509
+ {
510
+ "epoch": 0.09706774519716886,
511
+ "grad_norm": 0.431640625,
512
+ "learning_rate": 0.0007,
513
+ "loss": 2.3590816497802733,
514
+ "step": 1440
515
+ },
516
+ {
517
+ "epoch": 0.09841590832490732,
518
+ "grad_norm": 0.470703125,
519
+ "learning_rate": 0.0007,
520
+ "loss": 2.347888374328613,
521
+ "step": 1460
522
+ },
523
+ {
524
+ "epoch": 0.09976407145264576,
525
+ "grad_norm": 0.400390625,
526
+ "learning_rate": 0.0007,
527
+ "loss": 2.3355857849121096,
528
+ "step": 1480
529
+ },
530
+ {
531
+ "epoch": 0.10111223458038422,
532
+ "grad_norm": 0.4140625,
533
+ "learning_rate": 0.0007,
534
+ "loss": 2.3234769821166994,
535
+ "step": 1500
536
+ },
537
+ {
538
+ "epoch": 0.10246039770812268,
539
+ "grad_norm": 0.4375,
540
+ "learning_rate": 0.0007,
541
+ "loss": 2.327121543884277,
542
+ "step": 1520
543
+ },
544
+ {
545
+ "epoch": 0.10380856083586114,
546
+ "grad_norm": 0.408203125,
547
+ "learning_rate": 0.0007,
548
+ "loss": 2.3147512435913087,
549
+ "step": 1540
550
+ },
551
+ {
552
+ "epoch": 0.1051567239635996,
553
+ "grad_norm": 0.408203125,
554
+ "learning_rate": 0.0007,
555
+ "loss": 2.3023752212524413,
556
+ "step": 1560
557
+ },
558
+ {
559
+ "epoch": 0.10650488709133805,
560
+ "grad_norm": 0.408203125,
561
+ "learning_rate": 0.0007,
562
+ "loss": 2.3237445831298826,
563
+ "step": 1580
564
+ },
565
+ {
566
+ "epoch": 0.10785305021907651,
567
+ "grad_norm": 0.400390625,
568
+ "learning_rate": 0.0007,
569
+ "loss": 2.288845252990723,
570
+ "step": 1600
571
+ },
572
+ {
573
+ "epoch": 0.10920121334681497,
574
+ "grad_norm": 0.4140625,
575
+ "learning_rate": 0.0007,
576
+ "loss": 2.28785400390625,
577
+ "step": 1620
578
+ },
579
+ {
580
+ "epoch": 0.11054937647455342,
581
+ "grad_norm": 0.416015625,
582
+ "learning_rate": 0.0007,
583
+ "loss": 2.300959587097168,
584
+ "step": 1640
585
+ },
586
+ {
587
+ "epoch": 0.11189753960229187,
588
+ "grad_norm": 0.380859375,
589
+ "learning_rate": 0.0007,
590
+ "loss": 2.284528350830078,
591
+ "step": 1660
592
+ },
593
+ {
594
+ "epoch": 0.11324570273003033,
595
+ "grad_norm": 0.408203125,
596
+ "learning_rate": 0.0007,
597
+ "loss": 2.267934799194336,
598
+ "step": 1680
599
+ },
600
+ {
601
+ "epoch": 0.11459386585776879,
602
+ "grad_norm": 0.423828125,
603
+ "learning_rate": 0.0007,
604
+ "loss": 2.2802879333496096,
605
+ "step": 1700
606
+ },
607
+ {
608
+ "epoch": 0.11594202898550725,
609
+ "grad_norm": 0.4375,
610
+ "learning_rate": 0.0007,
611
+ "loss": 2.258160400390625,
612
+ "step": 1720
613
+ },
614
+ {
615
+ "epoch": 0.1172901921132457,
616
+ "grad_norm": 0.39453125,
617
+ "learning_rate": 0.0007,
618
+ "loss": 2.246730995178223,
619
+ "step": 1740
620
+ },
621
+ {
622
+ "epoch": 0.11863835524098416,
623
+ "grad_norm": 0.419921875,
624
+ "learning_rate": 0.0007,
625
+ "loss": 2.252003479003906,
626
+ "step": 1760
627
+ },
628
+ {
629
+ "epoch": 0.11998651836872262,
630
+ "grad_norm": 0.38671875,
631
+ "learning_rate": 0.0007,
632
+ "loss": 2.265170669555664,
633
+ "step": 1780
634
+ },
635
+ {
636
+ "epoch": 0.12133468149646107,
637
+ "grad_norm": 0.408203125,
638
+ "learning_rate": 0.0007,
639
+ "loss": 2.2610910415649412,
640
+ "step": 1800
641
+ },
642
+ {
643
+ "epoch": 0.12268284462419952,
644
+ "grad_norm": 0.41015625,
645
+ "learning_rate": 0.0007,
646
+ "loss": 2.2539751052856447,
647
+ "step": 1820
648
+ },
649
+ {
650
+ "epoch": 0.12403100775193798,
651
+ "grad_norm": 0.404296875,
652
+ "learning_rate": 0.0007,
653
+ "loss": 2.220688819885254,
654
+ "step": 1840
655
+ },
656
+ {
657
+ "epoch": 0.12537917087967643,
658
+ "grad_norm": 0.376953125,
659
+ "learning_rate": 0.0007,
660
+ "loss": 2.222416305541992,
661
+ "step": 1860
662
+ },
663
+ {
664
+ "epoch": 0.1267273340074149,
665
+ "grad_norm": 0.3828125,
666
+ "learning_rate": 0.0007,
667
+ "loss": 2.2135461807250976,
668
+ "step": 1880
669
+ },
670
+ {
671
+ "epoch": 0.12807549713515334,
672
+ "grad_norm": 0.39453125,
673
+ "learning_rate": 0.0007,
674
+ "loss": 2.2209815979003906,
675
+ "step": 1900
676
+ },
677
+ {
678
+ "epoch": 0.12942366026289182,
679
+ "grad_norm": 0.39453125,
680
+ "learning_rate": 0.0007,
681
+ "loss": 2.199201965332031,
682
+ "step": 1920
683
+ },
684
+ {
685
+ "epoch": 0.13077182339063026,
686
+ "grad_norm": 0.41015625,
687
+ "learning_rate": 0.0007,
688
+ "loss": 2.2096010208129884,
689
+ "step": 1940
690
+ },
691
+ {
692
+ "epoch": 0.13211998651836873,
693
+ "grad_norm": 0.396484375,
694
+ "learning_rate": 0.0007,
695
+ "loss": 2.1849014282226564,
696
+ "step": 1960
697
+ },
698
+ {
699
+ "epoch": 0.13346814964610718,
700
+ "grad_norm": 0.421875,
701
+ "learning_rate": 0.0007,
702
+ "loss": 2.2073038101196287,
703
+ "step": 1980
704
+ },
705
+ {
706
+ "epoch": 0.13481631277384565,
707
+ "grad_norm": 0.373046875,
708
+ "learning_rate": 0.0007,
709
+ "loss": 2.1818737030029296,
710
+ "step": 2000
711
+ },
712
+ {
713
+ "epoch": 0.1361644759015841,
714
+ "grad_norm": 0.384765625,
715
+ "learning_rate": 0.0007,
716
+ "loss": 2.1887689590454102,
717
+ "step": 2020
718
+ },
719
+ {
720
+ "epoch": 0.13751263902932254,
721
+ "grad_norm": 0.4296875,
722
+ "learning_rate": 0.0007,
723
+ "loss": 2.1714288711547853,
724
+ "step": 2040
725
+ },
726
+ {
727
+ "epoch": 0.138860802157061,
728
+ "grad_norm": 0.392578125,
729
+ "learning_rate": 0.0007,
730
+ "loss": 2.1905704498291017,
731
+ "step": 2060
732
+ },
733
+ {
734
+ "epoch": 0.14020896528479945,
735
+ "grad_norm": 0.388671875,
736
+ "learning_rate": 0.0007,
737
+ "loss": 2.168014717102051,
738
+ "step": 2080
739
+ },
740
+ {
741
+ "epoch": 0.14155712841253792,
742
+ "grad_norm": 0.37890625,
743
+ "learning_rate": 0.0007,
744
+ "loss": 2.157928466796875,
745
+ "step": 2100
746
+ },
747
+ {
748
+ "epoch": 0.14290529154027637,
749
+ "grad_norm": 0.376953125,
750
+ "learning_rate": 0.0007,
751
+ "loss": 2.1686134338378906,
752
+ "step": 2120
753
+ },
754
+ {
755
+ "epoch": 0.14425345466801484,
756
+ "grad_norm": 0.376953125,
757
+ "learning_rate": 0.0007,
758
+ "loss": 2.167608070373535,
759
+ "step": 2140
760
+ },
761
+ {
762
+ "epoch": 0.14560161779575329,
763
+ "grad_norm": 0.3828125,
764
+ "learning_rate": 0.0007,
765
+ "loss": 2.152691650390625,
766
+ "step": 2160
767
+ },
768
+ {
769
+ "epoch": 0.14694978092349173,
770
+ "grad_norm": 0.388671875,
771
+ "learning_rate": 0.0007,
772
+ "loss": 2.1620912551879883,
773
+ "step": 2180
774
+ },
775
+ {
776
+ "epoch": 0.1482979440512302,
777
+ "grad_norm": 0.388671875,
778
+ "learning_rate": 0.0007,
779
+ "loss": 2.1393999099731444,
780
+ "step": 2200
781
+ }
782
+ ],
783
+ "logging_steps": 20,
784
+ "max_steps": 2500,
785
+ "num_input_tokens_seen": 0,
786
+ "num_train_epochs": 1,
787
+ "save_steps": 100,
788
+ "stateful_callbacks": {
789
+ "TrainerControl": {
790
+ "args": {
791
+ "should_epoch_stop": false,
792
+ "should_evaluate": false,
793
+ "should_log": false,
794
+ "should_save": true,
795
+ "should_training_stop": false
796
+ },
797
+ "attributes": {}
798
+ }
799
+ },
800
+ "total_flos": 7097824365772800.0,
801
+ "train_batch_size": 64,
802
+ "trial_name": null,
803
+ "trial_params": null
804
+ }
zain/Activation/out/glu-waleed-100L_run/checkpoint-2200/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c75eb6d8bf9c953837c430a69915c00b95cc26c05d95e0643e1a66470a65c0cf
3
+ size 4920
zain/Activation/out/glu-waleed-100L_run/checkpoint-2300/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "waleed",
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-waleed-100L_run/checkpoint-2300/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:932663a0c0c8954a7b944cae0ffdf6309ea9bf41b0abd5511d0ca9d05bf24628
3
+ size 33967272
zain/Activation/out/glu-waleed-100L_run/checkpoint-2300/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:16c099f801c99ea453475232a460d46c1a15aa1b483f6540029f911e3595cba9
3
+ size 68504996
zain/Activation/out/glu-waleed-100L_run/checkpoint-2300/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-waleed-100L_run/checkpoint-2300/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b9bcfc4abc1cd7f19c4260bd52451981a53bf8db4a664a25c46ab7065952dd0
3
+ size 1064
zain/Activation/out/glu-waleed-100L_run/checkpoint-2300/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-waleed-100L_run/checkpoint-2300/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-waleed-100L_run/checkpoint-2300/trainer_state.json ADDED
@@ -0,0 +1,839 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.15503875968992248,
6
+ "eval_steps": 2498,
7
+ "global_step": 2300,
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": 2.125,
15
+ "learning_rate": 2.66e-05,
16
+ "loss": 8.299404907226563,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.3125,
22
+ "learning_rate": 5.46e-05,
23
+ "loss": 8.036402130126953,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 1.2734375,
29
+ "learning_rate": 8.259999999999999e-05,
30
+ "loss": 7.834580993652343,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 1.265625,
36
+ "learning_rate": 0.0001106,
37
+ "loss": 7.5719047546386715,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 1.1953125,
43
+ "learning_rate": 0.0001386,
44
+ "loss": 7.237163543701172,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 1.1484375,
50
+ "learning_rate": 0.00016659999999999998,
51
+ "loss": 6.864958190917969,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 1.015625,
57
+ "learning_rate": 0.00019460000000000001,
58
+ "loss": 6.509477996826172,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 1.0234375,
64
+ "learning_rate": 0.0002226,
65
+ "loss": 6.204301452636718,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 1.0859375,
71
+ "learning_rate": 0.00025059999999999997,
72
+ "loss": 5.940593719482422,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 1.2109375,
78
+ "learning_rate": 0.0002786,
79
+ "loss": 5.676498794555664,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.01482979440512302,
84
+ "grad_norm": 0.8046875,
85
+ "learning_rate": 0.00030659999999999997,
86
+ "loss": 5.412863540649414,
87
+ "step": 220
88
+ },
89
+ {
90
+ "epoch": 0.016177957532861477,
91
+ "grad_norm": 0.71875,
92
+ "learning_rate": 0.0003346,
93
+ "loss": 5.1301216125488285,
94
+ "step": 240
95
+ },
96
+ {
97
+ "epoch": 0.01752612066059993,
98
+ "grad_norm": 1.046875,
99
+ "learning_rate": 0.00036260000000000003,
100
+ "loss": 4.88237190246582,
101
+ "step": 260
102
+ },
103
+ {
104
+ "epoch": 0.01887428378833839,
105
+ "grad_norm": 0.75,
106
+ "learning_rate": 0.0003906,
107
+ "loss": 4.643666076660156,
108
+ "step": 280
109
+ },
110
+ {
111
+ "epoch": 0.020222446916076844,
112
+ "grad_norm": 0.83984375,
113
+ "learning_rate": 0.0004186,
114
+ "loss": 4.441861343383789,
115
+ "step": 300
116
+ },
117
+ {
118
+ "epoch": 0.021570610043815303,
119
+ "grad_norm": 0.8125,
120
+ "learning_rate": 0.0004466,
121
+ "loss": 4.25420036315918,
122
+ "step": 320
123
+ },
124
+ {
125
+ "epoch": 0.022918773171553757,
126
+ "grad_norm": 0.62109375,
127
+ "learning_rate": 0.00047460000000000004,
128
+ "loss": 4.097484970092774,
129
+ "step": 340
130
+ },
131
+ {
132
+ "epoch": 0.024266936299292215,
133
+ "grad_norm": 0.6015625,
134
+ "learning_rate": 0.0005026,
135
+ "loss": 3.9764480590820312,
136
+ "step": 360
137
+ },
138
+ {
139
+ "epoch": 0.02561509942703067,
140
+ "grad_norm": 0.9140625,
141
+ "learning_rate": 0.0005306,
142
+ "loss": 3.854802703857422,
143
+ "step": 380
144
+ },
145
+ {
146
+ "epoch": 0.026963262554769128,
147
+ "grad_norm": 0.56640625,
148
+ "learning_rate": 0.0005586,
149
+ "loss": 3.7329875946044924,
150
+ "step": 400
151
+ },
152
+ {
153
+ "epoch": 0.028311425682507583,
154
+ "grad_norm": 0.53515625,
155
+ "learning_rate": 0.0005866,
156
+ "loss": 3.6484100341796877,
157
+ "step": 420
158
+ },
159
+ {
160
+ "epoch": 0.02965958881024604,
161
+ "grad_norm": 0.5390625,
162
+ "learning_rate": 0.0006146,
163
+ "loss": 3.595387268066406,
164
+ "step": 440
165
+ },
166
+ {
167
+ "epoch": 0.031007751937984496,
168
+ "grad_norm": 0.55859375,
169
+ "learning_rate": 0.0006426,
170
+ "loss": 3.498302459716797,
171
+ "step": 460
172
+ },
173
+ {
174
+ "epoch": 0.032355915065722954,
175
+ "grad_norm": 0.5859375,
176
+ "learning_rate": 0.0006705999999999999,
177
+ "loss": 3.4313674926757813,
178
+ "step": 480
179
+ },
180
+ {
181
+ "epoch": 0.03370407819346141,
182
+ "grad_norm": 0.61328125,
183
+ "learning_rate": 0.0006986,
184
+ "loss": 3.3613121032714846,
185
+ "step": 500
186
+ },
187
+ {
188
+ "epoch": 0.03505224132119986,
189
+ "grad_norm": 0.474609375,
190
+ "learning_rate": 0.0007,
191
+ "loss": 3.3039566040039063,
192
+ "step": 520
193
+ },
194
+ {
195
+ "epoch": 0.03640040444893832,
196
+ "grad_norm": 0.578125,
197
+ "learning_rate": 0.0007,
198
+ "loss": 3.2328567504882812,
199
+ "step": 540
200
+ },
201
+ {
202
+ "epoch": 0.03774856757667678,
203
+ "grad_norm": 0.64453125,
204
+ "learning_rate": 0.0007,
205
+ "loss": 3.187940788269043,
206
+ "step": 560
207
+ },
208
+ {
209
+ "epoch": 0.03909673070441524,
210
+ "grad_norm": 0.462890625,
211
+ "learning_rate": 0.0007,
212
+ "loss": 3.133651542663574,
213
+ "step": 580
214
+ },
215
+ {
216
+ "epoch": 0.04044489383215369,
217
+ "grad_norm": 0.5,
218
+ "learning_rate": 0.0007,
219
+ "loss": 3.1042255401611327,
220
+ "step": 600
221
+ },
222
+ {
223
+ "epoch": 0.04179305695989215,
224
+ "grad_norm": 0.578125,
225
+ "learning_rate": 0.0007,
226
+ "loss": 3.04781551361084,
227
+ "step": 620
228
+ },
229
+ {
230
+ "epoch": 0.043141220087630605,
231
+ "grad_norm": 0.453125,
232
+ "learning_rate": 0.0007,
233
+ "loss": 3.0298320770263674,
234
+ "step": 640
235
+ },
236
+ {
237
+ "epoch": 0.044489383215369056,
238
+ "grad_norm": 0.46875,
239
+ "learning_rate": 0.0007,
240
+ "loss": 2.9790300369262694,
241
+ "step": 660
242
+ },
243
+ {
244
+ "epoch": 0.045837546343107514,
245
+ "grad_norm": 0.46875,
246
+ "learning_rate": 0.0007,
247
+ "loss": 2.9566295623779295,
248
+ "step": 680
249
+ },
250
+ {
251
+ "epoch": 0.04718570947084597,
252
+ "grad_norm": 0.439453125,
253
+ "learning_rate": 0.0007,
254
+ "loss": 2.916979217529297,
255
+ "step": 700
256
+ },
257
+ {
258
+ "epoch": 0.04853387259858443,
259
+ "grad_norm": 0.41796875,
260
+ "learning_rate": 0.0007,
261
+ "loss": 2.8888761520385744,
262
+ "step": 720
263
+ },
264
+ {
265
+ "epoch": 0.04988203572632288,
266
+ "grad_norm": 0.4765625,
267
+ "learning_rate": 0.0007,
268
+ "loss": 2.855138969421387,
269
+ "step": 740
270
+ },
271
+ {
272
+ "epoch": 0.05123019885406134,
273
+ "grad_norm": 0.46875,
274
+ "learning_rate": 0.0007,
275
+ "loss": 2.841838836669922,
276
+ "step": 760
277
+ },
278
+ {
279
+ "epoch": 0.0525783619817998,
280
+ "grad_norm": 0.43359375,
281
+ "learning_rate": 0.0007,
282
+ "loss": 2.8384235382080076,
283
+ "step": 780
284
+ },
285
+ {
286
+ "epoch": 0.053926525109538256,
287
+ "grad_norm": 0.447265625,
288
+ "learning_rate": 0.0007,
289
+ "loss": 2.7947757720947264,
290
+ "step": 800
291
+ },
292
+ {
293
+ "epoch": 0.05527468823727671,
294
+ "grad_norm": 0.478515625,
295
+ "learning_rate": 0.0007,
296
+ "loss": 2.7847801208496095,
297
+ "step": 820
298
+ },
299
+ {
300
+ "epoch": 0.056622851365015166,
301
+ "grad_norm": 0.44140625,
302
+ "learning_rate": 0.0007,
303
+ "loss": 2.7601039886474608,
304
+ "step": 840
305
+ },
306
+ {
307
+ "epoch": 0.057971014492753624,
308
+ "grad_norm": 0.484375,
309
+ "learning_rate": 0.0007,
310
+ "loss": 2.716906929016113,
311
+ "step": 860
312
+ },
313
+ {
314
+ "epoch": 0.05931917762049208,
315
+ "grad_norm": 0.439453125,
316
+ "learning_rate": 0.0007,
317
+ "loss": 2.6968852996826174,
318
+ "step": 880
319
+ },
320
+ {
321
+ "epoch": 0.06066734074823053,
322
+ "grad_norm": 0.447265625,
323
+ "learning_rate": 0.0007,
324
+ "loss": 2.6847219467163086,
325
+ "step": 900
326
+ },
327
+ {
328
+ "epoch": 0.06201550387596899,
329
+ "grad_norm": 0.435546875,
330
+ "learning_rate": 0.0007,
331
+ "loss": 2.6599369049072266,
332
+ "step": 920
333
+ },
334
+ {
335
+ "epoch": 0.06336366700370745,
336
+ "grad_norm": 0.455078125,
337
+ "learning_rate": 0.0007,
338
+ "loss": 2.6465431213378907,
339
+ "step": 940
340
+ },
341
+ {
342
+ "epoch": 0.06471183013144591,
343
+ "grad_norm": 0.455078125,
344
+ "learning_rate": 0.0007,
345
+ "loss": 2.623439979553223,
346
+ "step": 960
347
+ },
348
+ {
349
+ "epoch": 0.06605999325918437,
350
+ "grad_norm": 0.462890625,
351
+ "learning_rate": 0.0007,
352
+ "loss": 2.609505271911621,
353
+ "step": 980
354
+ },
355
+ {
356
+ "epoch": 0.06740815638692282,
357
+ "grad_norm": 0.48828125,
358
+ "learning_rate": 0.0007,
359
+ "loss": 2.610812187194824,
360
+ "step": 1000
361
+ },
362
+ {
363
+ "epoch": 0.06875631951466127,
364
+ "grad_norm": 0.423828125,
365
+ "learning_rate": 0.0007,
366
+ "loss": 2.581887054443359,
367
+ "step": 1020
368
+ },
369
+ {
370
+ "epoch": 0.07010448264239973,
371
+ "grad_norm": 0.4453125,
372
+ "learning_rate": 0.0007,
373
+ "loss": 2.5679080963134764,
374
+ "step": 1040
375
+ },
376
+ {
377
+ "epoch": 0.07145264577013818,
378
+ "grad_norm": 0.447265625,
379
+ "learning_rate": 0.0007,
380
+ "loss": 2.55521240234375,
381
+ "step": 1060
382
+ },
383
+ {
384
+ "epoch": 0.07280080889787664,
385
+ "grad_norm": 0.4453125,
386
+ "learning_rate": 0.0007,
387
+ "loss": 2.536572074890137,
388
+ "step": 1080
389
+ },
390
+ {
391
+ "epoch": 0.0741489720256151,
392
+ "grad_norm": 0.431640625,
393
+ "learning_rate": 0.0007,
394
+ "loss": 2.5341920852661133,
395
+ "step": 1100
396
+ },
397
+ {
398
+ "epoch": 0.07549713515335356,
399
+ "grad_norm": 0.427734375,
400
+ "learning_rate": 0.0007,
401
+ "loss": 2.4994813919067385,
402
+ "step": 1120
403
+ },
404
+ {
405
+ "epoch": 0.07684529828109202,
406
+ "grad_norm": 0.421875,
407
+ "learning_rate": 0.0007,
408
+ "loss": 2.4886734008789064,
409
+ "step": 1140
410
+ },
411
+ {
412
+ "epoch": 0.07819346140883048,
413
+ "grad_norm": 0.419921875,
414
+ "learning_rate": 0.0007,
415
+ "loss": 2.4889175415039064,
416
+ "step": 1160
417
+ },
418
+ {
419
+ "epoch": 0.07954162453656892,
420
+ "grad_norm": 0.4140625,
421
+ "learning_rate": 0.0007,
422
+ "loss": 2.484502410888672,
423
+ "step": 1180
424
+ },
425
+ {
426
+ "epoch": 0.08088978766430738,
427
+ "grad_norm": 0.51171875,
428
+ "learning_rate": 0.0007,
429
+ "loss": 2.476539421081543,
430
+ "step": 1200
431
+ },
432
+ {
433
+ "epoch": 0.08223795079204584,
434
+ "grad_norm": 0.41796875,
435
+ "learning_rate": 0.0007,
436
+ "loss": 2.463785171508789,
437
+ "step": 1220
438
+ },
439
+ {
440
+ "epoch": 0.0835861139197843,
441
+ "grad_norm": 0.390625,
442
+ "learning_rate": 0.0007,
443
+ "loss": 2.4509374618530275,
444
+ "step": 1240
445
+ },
446
+ {
447
+ "epoch": 0.08493427704752275,
448
+ "grad_norm": 0.4453125,
449
+ "learning_rate": 0.0007,
450
+ "loss": 2.4406501770019533,
451
+ "step": 1260
452
+ },
453
+ {
454
+ "epoch": 0.08628244017526121,
455
+ "grad_norm": 0.41796875,
456
+ "learning_rate": 0.0007,
457
+ "loss": 2.423234748840332,
458
+ "step": 1280
459
+ },
460
+ {
461
+ "epoch": 0.08763060330299967,
462
+ "grad_norm": 0.4453125,
463
+ "learning_rate": 0.0007,
464
+ "loss": 2.415129280090332,
465
+ "step": 1300
466
+ },
467
+ {
468
+ "epoch": 0.08897876643073811,
469
+ "grad_norm": 0.41015625,
470
+ "learning_rate": 0.0007,
471
+ "loss": 2.4075996398925783,
472
+ "step": 1320
473
+ },
474
+ {
475
+ "epoch": 0.09032692955847657,
476
+ "grad_norm": 0.412109375,
477
+ "learning_rate": 0.0007,
478
+ "loss": 2.396971321105957,
479
+ "step": 1340
480
+ },
481
+ {
482
+ "epoch": 0.09167509268621503,
483
+ "grad_norm": 0.431640625,
484
+ "learning_rate": 0.0007,
485
+ "loss": 2.398550605773926,
486
+ "step": 1360
487
+ },
488
+ {
489
+ "epoch": 0.09302325581395349,
490
+ "grad_norm": 0.40234375,
491
+ "learning_rate": 0.0007,
492
+ "loss": 2.3772783279418945,
493
+ "step": 1380
494
+ },
495
+ {
496
+ "epoch": 0.09437141894169195,
497
+ "grad_norm": 0.40625,
498
+ "learning_rate": 0.0007,
499
+ "loss": 2.3700124740600588,
500
+ "step": 1400
501
+ },
502
+ {
503
+ "epoch": 0.0957195820694304,
504
+ "grad_norm": 0.421875,
505
+ "learning_rate": 0.0007,
506
+ "loss": 2.3814950942993165,
507
+ "step": 1420
508
+ },
509
+ {
510
+ "epoch": 0.09706774519716886,
511
+ "grad_norm": 0.431640625,
512
+ "learning_rate": 0.0007,
513
+ "loss": 2.3590816497802733,
514
+ "step": 1440
515
+ },
516
+ {
517
+ "epoch": 0.09841590832490732,
518
+ "grad_norm": 0.470703125,
519
+ "learning_rate": 0.0007,
520
+ "loss": 2.347888374328613,
521
+ "step": 1460
522
+ },
523
+ {
524
+ "epoch": 0.09976407145264576,
525
+ "grad_norm": 0.400390625,
526
+ "learning_rate": 0.0007,
527
+ "loss": 2.3355857849121096,
528
+ "step": 1480
529
+ },
530
+ {
531
+ "epoch": 0.10111223458038422,
532
+ "grad_norm": 0.4140625,
533
+ "learning_rate": 0.0007,
534
+ "loss": 2.3234769821166994,
535
+ "step": 1500
536
+ },
537
+ {
538
+ "epoch": 0.10246039770812268,
539
+ "grad_norm": 0.4375,
540
+ "learning_rate": 0.0007,
541
+ "loss": 2.327121543884277,
542
+ "step": 1520
543
+ },
544
+ {
545
+ "epoch": 0.10380856083586114,
546
+ "grad_norm": 0.408203125,
547
+ "learning_rate": 0.0007,
548
+ "loss": 2.3147512435913087,
549
+ "step": 1540
550
+ },
551
+ {
552
+ "epoch": 0.1051567239635996,
553
+ "grad_norm": 0.408203125,
554
+ "learning_rate": 0.0007,
555
+ "loss": 2.3023752212524413,
556
+ "step": 1560
557
+ },
558
+ {
559
+ "epoch": 0.10650488709133805,
560
+ "grad_norm": 0.408203125,
561
+ "learning_rate": 0.0007,
562
+ "loss": 2.3237445831298826,
563
+ "step": 1580
564
+ },
565
+ {
566
+ "epoch": 0.10785305021907651,
567
+ "grad_norm": 0.400390625,
568
+ "learning_rate": 0.0007,
569
+ "loss": 2.288845252990723,
570
+ "step": 1600
571
+ },
572
+ {
573
+ "epoch": 0.10920121334681497,
574
+ "grad_norm": 0.4140625,
575
+ "learning_rate": 0.0007,
576
+ "loss": 2.28785400390625,
577
+ "step": 1620
578
+ },
579
+ {
580
+ "epoch": 0.11054937647455342,
581
+ "grad_norm": 0.416015625,
582
+ "learning_rate": 0.0007,
583
+ "loss": 2.300959587097168,
584
+ "step": 1640
585
+ },
586
+ {
587
+ "epoch": 0.11189753960229187,
588
+ "grad_norm": 0.380859375,
589
+ "learning_rate": 0.0007,
590
+ "loss": 2.284528350830078,
591
+ "step": 1660
592
+ },
593
+ {
594
+ "epoch": 0.11324570273003033,
595
+ "grad_norm": 0.408203125,
596
+ "learning_rate": 0.0007,
597
+ "loss": 2.267934799194336,
598
+ "step": 1680
599
+ },
600
+ {
601
+ "epoch": 0.11459386585776879,
602
+ "grad_norm": 0.423828125,
603
+ "learning_rate": 0.0007,
604
+ "loss": 2.2802879333496096,
605
+ "step": 1700
606
+ },
607
+ {
608
+ "epoch": 0.11594202898550725,
609
+ "grad_norm": 0.4375,
610
+ "learning_rate": 0.0007,
611
+ "loss": 2.258160400390625,
612
+ "step": 1720
613
+ },
614
+ {
615
+ "epoch": 0.1172901921132457,
616
+ "grad_norm": 0.39453125,
617
+ "learning_rate": 0.0007,
618
+ "loss": 2.246730995178223,
619
+ "step": 1740
620
+ },
621
+ {
622
+ "epoch": 0.11863835524098416,
623
+ "grad_norm": 0.419921875,
624
+ "learning_rate": 0.0007,
625
+ "loss": 2.252003479003906,
626
+ "step": 1760
627
+ },
628
+ {
629
+ "epoch": 0.11998651836872262,
630
+ "grad_norm": 0.38671875,
631
+ "learning_rate": 0.0007,
632
+ "loss": 2.265170669555664,
633
+ "step": 1780
634
+ },
635
+ {
636
+ "epoch": 0.12133468149646107,
637
+ "grad_norm": 0.408203125,
638
+ "learning_rate": 0.0007,
639
+ "loss": 2.2610910415649412,
640
+ "step": 1800
641
+ },
642
+ {
643
+ "epoch": 0.12268284462419952,
644
+ "grad_norm": 0.41015625,
645
+ "learning_rate": 0.0007,
646
+ "loss": 2.2539751052856447,
647
+ "step": 1820
648
+ },
649
+ {
650
+ "epoch": 0.12403100775193798,
651
+ "grad_norm": 0.404296875,
652
+ "learning_rate": 0.0007,
653
+ "loss": 2.220688819885254,
654
+ "step": 1840
655
+ },
656
+ {
657
+ "epoch": 0.12537917087967643,
658
+ "grad_norm": 0.376953125,
659
+ "learning_rate": 0.0007,
660
+ "loss": 2.222416305541992,
661
+ "step": 1860
662
+ },
663
+ {
664
+ "epoch": 0.1267273340074149,
665
+ "grad_norm": 0.3828125,
666
+ "learning_rate": 0.0007,
667
+ "loss": 2.2135461807250976,
668
+ "step": 1880
669
+ },
670
+ {
671
+ "epoch": 0.12807549713515334,
672
+ "grad_norm": 0.39453125,
673
+ "learning_rate": 0.0007,
674
+ "loss": 2.2209815979003906,
675
+ "step": 1900
676
+ },
677
+ {
678
+ "epoch": 0.12942366026289182,
679
+ "grad_norm": 0.39453125,
680
+ "learning_rate": 0.0007,
681
+ "loss": 2.199201965332031,
682
+ "step": 1920
683
+ },
684
+ {
685
+ "epoch": 0.13077182339063026,
686
+ "grad_norm": 0.41015625,
687
+ "learning_rate": 0.0007,
688
+ "loss": 2.2096010208129884,
689
+ "step": 1940
690
+ },
691
+ {
692
+ "epoch": 0.13211998651836873,
693
+ "grad_norm": 0.396484375,
694
+ "learning_rate": 0.0007,
695
+ "loss": 2.1849014282226564,
696
+ "step": 1960
697
+ },
698
+ {
699
+ "epoch": 0.13346814964610718,
700
+ "grad_norm": 0.421875,
701
+ "learning_rate": 0.0007,
702
+ "loss": 2.2073038101196287,
703
+ "step": 1980
704
+ },
705
+ {
706
+ "epoch": 0.13481631277384565,
707
+ "grad_norm": 0.373046875,
708
+ "learning_rate": 0.0007,
709
+ "loss": 2.1818737030029296,
710
+ "step": 2000
711
+ },
712
+ {
713
+ "epoch": 0.1361644759015841,
714
+ "grad_norm": 0.384765625,
715
+ "learning_rate": 0.0007,
716
+ "loss": 2.1887689590454102,
717
+ "step": 2020
718
+ },
719
+ {
720
+ "epoch": 0.13751263902932254,
721
+ "grad_norm": 0.4296875,
722
+ "learning_rate": 0.0007,
723
+ "loss": 2.1714288711547853,
724
+ "step": 2040
725
+ },
726
+ {
727
+ "epoch": 0.138860802157061,
728
+ "grad_norm": 0.392578125,
729
+ "learning_rate": 0.0007,
730
+ "loss": 2.1905704498291017,
731
+ "step": 2060
732
+ },
733
+ {
734
+ "epoch": 0.14020896528479945,
735
+ "grad_norm": 0.388671875,
736
+ "learning_rate": 0.0007,
737
+ "loss": 2.168014717102051,
738
+ "step": 2080
739
+ },
740
+ {
741
+ "epoch": 0.14155712841253792,
742
+ "grad_norm": 0.37890625,
743
+ "learning_rate": 0.0007,
744
+ "loss": 2.157928466796875,
745
+ "step": 2100
746
+ },
747
+ {
748
+ "epoch": 0.14290529154027637,
749
+ "grad_norm": 0.376953125,
750
+ "learning_rate": 0.0007,
751
+ "loss": 2.1686134338378906,
752
+ "step": 2120
753
+ },
754
+ {
755
+ "epoch": 0.14425345466801484,
756
+ "grad_norm": 0.376953125,
757
+ "learning_rate": 0.0007,
758
+ "loss": 2.167608070373535,
759
+ "step": 2140
760
+ },
761
+ {
762
+ "epoch": 0.14560161779575329,
763
+ "grad_norm": 0.3828125,
764
+ "learning_rate": 0.0007,
765
+ "loss": 2.152691650390625,
766
+ "step": 2160
767
+ },
768
+ {
769
+ "epoch": 0.14694978092349173,
770
+ "grad_norm": 0.388671875,
771
+ "learning_rate": 0.0007,
772
+ "loss": 2.1620912551879883,
773
+ "step": 2180
774
+ },
775
+ {
776
+ "epoch": 0.1482979440512302,
777
+ "grad_norm": 0.388671875,
778
+ "learning_rate": 0.0007,
779
+ "loss": 2.1393999099731444,
780
+ "step": 2200
781
+ },
782
+ {
783
+ "epoch": 0.14964610717896865,
784
+ "grad_norm": 0.375,
785
+ "learning_rate": 0.0007,
786
+ "loss": 2.14599494934082,
787
+ "step": 2220
788
+ },
789
+ {
790
+ "epoch": 0.15099427030670712,
791
+ "grad_norm": 0.3984375,
792
+ "learning_rate": 0.0007,
793
+ "loss": 2.1544084548950195,
794
+ "step": 2240
795
+ },
796
+ {
797
+ "epoch": 0.15234243343444556,
798
+ "grad_norm": 0.3984375,
799
+ "learning_rate": 0.0007,
800
+ "loss": 2.132464599609375,
801
+ "step": 2260
802
+ },
803
+ {
804
+ "epoch": 0.15369059656218403,
805
+ "grad_norm": 0.373046875,
806
+ "learning_rate": 0.0007,
807
+ "loss": 2.143385314941406,
808
+ "step": 2280
809
+ },
810
+ {
811
+ "epoch": 0.15503875968992248,
812
+ "grad_norm": 0.376953125,
813
+ "learning_rate": 0.0007,
814
+ "loss": 2.1241661071777345,
815
+ "step": 2300
816
+ }
817
+ ],
818
+ "logging_steps": 20,
819
+ "max_steps": 2500,
820
+ "num_input_tokens_seen": 0,
821
+ "num_train_epochs": 1,
822
+ "save_steps": 100,
823
+ "stateful_callbacks": {
824
+ "TrainerControl": {
825
+ "args": {
826
+ "should_epoch_stop": false,
827
+ "should_evaluate": false,
828
+ "should_log": false,
829
+ "should_save": true,
830
+ "should_training_stop": false
831
+ },
832
+ "attributes": {}
833
+ }
834
+ },
835
+ "total_flos": 7420452746035200.0,
836
+ "train_batch_size": 64,
837
+ "trial_name": null,
838
+ "trial_params": null
839
+ }
zain/Activation/out/glu-waleed-100L_run/checkpoint-2300/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c75eb6d8bf9c953837c430a69915c00b95cc26c05d95e0643e1a66470a65c0cf
3
+ size 4920
zain/Activation/out/glu-waleed-100L_run/checkpoint-2400/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "waleed",
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-waleed-100L_run/checkpoint-2400/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:447afac86d6d0567b30e73d1482055ff811955d406db262ce3970b12e4402c80
3
+ size 33967272
zain/Activation/out/glu-waleed-100L_run/checkpoint-2400/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:718bd811fece137f42ffa50af3a11ec486f0da85cb5d2a7527fea5c34fafb78c
3
+ size 68504996
zain/Activation/out/glu-waleed-100L_run/checkpoint-2400/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-waleed-100L_run/checkpoint-2400/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cffe3454608d7821f042822654b62bd9020d78872ebdecde4f22f94320596378
3
+ size 1064
zain/Activation/out/glu-waleed-100L_run/checkpoint-2400/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-waleed-100L_run/checkpoint-2400/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-waleed-100L_run/checkpoint-2400/trainer_state.json ADDED
@@ -0,0 +1,874 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.16177957532861476,
6
+ "eval_steps": 2498,
7
+ "global_step": 2400,
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": 2.125,
15
+ "learning_rate": 2.66e-05,
16
+ "loss": 8.299404907226563,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.3125,
22
+ "learning_rate": 5.46e-05,
23
+ "loss": 8.036402130126953,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 1.2734375,
29
+ "learning_rate": 8.259999999999999e-05,
30
+ "loss": 7.834580993652343,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 1.265625,
36
+ "learning_rate": 0.0001106,
37
+ "loss": 7.5719047546386715,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 1.1953125,
43
+ "learning_rate": 0.0001386,
44
+ "loss": 7.237163543701172,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 1.1484375,
50
+ "learning_rate": 0.00016659999999999998,
51
+ "loss": 6.864958190917969,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 1.015625,
57
+ "learning_rate": 0.00019460000000000001,
58
+ "loss": 6.509477996826172,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 1.0234375,
64
+ "learning_rate": 0.0002226,
65
+ "loss": 6.204301452636718,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 1.0859375,
71
+ "learning_rate": 0.00025059999999999997,
72
+ "loss": 5.940593719482422,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 1.2109375,
78
+ "learning_rate": 0.0002786,
79
+ "loss": 5.676498794555664,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.01482979440512302,
84
+ "grad_norm": 0.8046875,
85
+ "learning_rate": 0.00030659999999999997,
86
+ "loss": 5.412863540649414,
87
+ "step": 220
88
+ },
89
+ {
90
+ "epoch": 0.016177957532861477,
91
+ "grad_norm": 0.71875,
92
+ "learning_rate": 0.0003346,
93
+ "loss": 5.1301216125488285,
94
+ "step": 240
95
+ },
96
+ {
97
+ "epoch": 0.01752612066059993,
98
+ "grad_norm": 1.046875,
99
+ "learning_rate": 0.00036260000000000003,
100
+ "loss": 4.88237190246582,
101
+ "step": 260
102
+ },
103
+ {
104
+ "epoch": 0.01887428378833839,
105
+ "grad_norm": 0.75,
106
+ "learning_rate": 0.0003906,
107
+ "loss": 4.643666076660156,
108
+ "step": 280
109
+ },
110
+ {
111
+ "epoch": 0.020222446916076844,
112
+ "grad_norm": 0.83984375,
113
+ "learning_rate": 0.0004186,
114
+ "loss": 4.441861343383789,
115
+ "step": 300
116
+ },
117
+ {
118
+ "epoch": 0.021570610043815303,
119
+ "grad_norm": 0.8125,
120
+ "learning_rate": 0.0004466,
121
+ "loss": 4.25420036315918,
122
+ "step": 320
123
+ },
124
+ {
125
+ "epoch": 0.022918773171553757,
126
+ "grad_norm": 0.62109375,
127
+ "learning_rate": 0.00047460000000000004,
128
+ "loss": 4.097484970092774,
129
+ "step": 340
130
+ },
131
+ {
132
+ "epoch": 0.024266936299292215,
133
+ "grad_norm": 0.6015625,
134
+ "learning_rate": 0.0005026,
135
+ "loss": 3.9764480590820312,
136
+ "step": 360
137
+ },
138
+ {
139
+ "epoch": 0.02561509942703067,
140
+ "grad_norm": 0.9140625,
141
+ "learning_rate": 0.0005306,
142
+ "loss": 3.854802703857422,
143
+ "step": 380
144
+ },
145
+ {
146
+ "epoch": 0.026963262554769128,
147
+ "grad_norm": 0.56640625,
148
+ "learning_rate": 0.0005586,
149
+ "loss": 3.7329875946044924,
150
+ "step": 400
151
+ },
152
+ {
153
+ "epoch": 0.028311425682507583,
154
+ "grad_norm": 0.53515625,
155
+ "learning_rate": 0.0005866,
156
+ "loss": 3.6484100341796877,
157
+ "step": 420
158
+ },
159
+ {
160
+ "epoch": 0.02965958881024604,
161
+ "grad_norm": 0.5390625,
162
+ "learning_rate": 0.0006146,
163
+ "loss": 3.595387268066406,
164
+ "step": 440
165
+ },
166
+ {
167
+ "epoch": 0.031007751937984496,
168
+ "grad_norm": 0.55859375,
169
+ "learning_rate": 0.0006426,
170
+ "loss": 3.498302459716797,
171
+ "step": 460
172
+ },
173
+ {
174
+ "epoch": 0.032355915065722954,
175
+ "grad_norm": 0.5859375,
176
+ "learning_rate": 0.0006705999999999999,
177
+ "loss": 3.4313674926757813,
178
+ "step": 480
179
+ },
180
+ {
181
+ "epoch": 0.03370407819346141,
182
+ "grad_norm": 0.61328125,
183
+ "learning_rate": 0.0006986,
184
+ "loss": 3.3613121032714846,
185
+ "step": 500
186
+ },
187
+ {
188
+ "epoch": 0.03505224132119986,
189
+ "grad_norm": 0.474609375,
190
+ "learning_rate": 0.0007,
191
+ "loss": 3.3039566040039063,
192
+ "step": 520
193
+ },
194
+ {
195
+ "epoch": 0.03640040444893832,
196
+ "grad_norm": 0.578125,
197
+ "learning_rate": 0.0007,
198
+ "loss": 3.2328567504882812,
199
+ "step": 540
200
+ },
201
+ {
202
+ "epoch": 0.03774856757667678,
203
+ "grad_norm": 0.64453125,
204
+ "learning_rate": 0.0007,
205
+ "loss": 3.187940788269043,
206
+ "step": 560
207
+ },
208
+ {
209
+ "epoch": 0.03909673070441524,
210
+ "grad_norm": 0.462890625,
211
+ "learning_rate": 0.0007,
212
+ "loss": 3.133651542663574,
213
+ "step": 580
214
+ },
215
+ {
216
+ "epoch": 0.04044489383215369,
217
+ "grad_norm": 0.5,
218
+ "learning_rate": 0.0007,
219
+ "loss": 3.1042255401611327,
220
+ "step": 600
221
+ },
222
+ {
223
+ "epoch": 0.04179305695989215,
224
+ "grad_norm": 0.578125,
225
+ "learning_rate": 0.0007,
226
+ "loss": 3.04781551361084,
227
+ "step": 620
228
+ },
229
+ {
230
+ "epoch": 0.043141220087630605,
231
+ "grad_norm": 0.453125,
232
+ "learning_rate": 0.0007,
233
+ "loss": 3.0298320770263674,
234
+ "step": 640
235
+ },
236
+ {
237
+ "epoch": 0.044489383215369056,
238
+ "grad_norm": 0.46875,
239
+ "learning_rate": 0.0007,
240
+ "loss": 2.9790300369262694,
241
+ "step": 660
242
+ },
243
+ {
244
+ "epoch": 0.045837546343107514,
245
+ "grad_norm": 0.46875,
246
+ "learning_rate": 0.0007,
247
+ "loss": 2.9566295623779295,
248
+ "step": 680
249
+ },
250
+ {
251
+ "epoch": 0.04718570947084597,
252
+ "grad_norm": 0.439453125,
253
+ "learning_rate": 0.0007,
254
+ "loss": 2.916979217529297,
255
+ "step": 700
256
+ },
257
+ {
258
+ "epoch": 0.04853387259858443,
259
+ "grad_norm": 0.41796875,
260
+ "learning_rate": 0.0007,
261
+ "loss": 2.8888761520385744,
262
+ "step": 720
263
+ },
264
+ {
265
+ "epoch": 0.04988203572632288,
266
+ "grad_norm": 0.4765625,
267
+ "learning_rate": 0.0007,
268
+ "loss": 2.855138969421387,
269
+ "step": 740
270
+ },
271
+ {
272
+ "epoch": 0.05123019885406134,
273
+ "grad_norm": 0.46875,
274
+ "learning_rate": 0.0007,
275
+ "loss": 2.841838836669922,
276
+ "step": 760
277
+ },
278
+ {
279
+ "epoch": 0.0525783619817998,
280
+ "grad_norm": 0.43359375,
281
+ "learning_rate": 0.0007,
282
+ "loss": 2.8384235382080076,
283
+ "step": 780
284
+ },
285
+ {
286
+ "epoch": 0.053926525109538256,
287
+ "grad_norm": 0.447265625,
288
+ "learning_rate": 0.0007,
289
+ "loss": 2.7947757720947264,
290
+ "step": 800
291
+ },
292
+ {
293
+ "epoch": 0.05527468823727671,
294
+ "grad_norm": 0.478515625,
295
+ "learning_rate": 0.0007,
296
+ "loss": 2.7847801208496095,
297
+ "step": 820
298
+ },
299
+ {
300
+ "epoch": 0.056622851365015166,
301
+ "grad_norm": 0.44140625,
302
+ "learning_rate": 0.0007,
303
+ "loss": 2.7601039886474608,
304
+ "step": 840
305
+ },
306
+ {
307
+ "epoch": 0.057971014492753624,
308
+ "grad_norm": 0.484375,
309
+ "learning_rate": 0.0007,
310
+ "loss": 2.716906929016113,
311
+ "step": 860
312
+ },
313
+ {
314
+ "epoch": 0.05931917762049208,
315
+ "grad_norm": 0.439453125,
316
+ "learning_rate": 0.0007,
317
+ "loss": 2.6968852996826174,
318
+ "step": 880
319
+ },
320
+ {
321
+ "epoch": 0.06066734074823053,
322
+ "grad_norm": 0.447265625,
323
+ "learning_rate": 0.0007,
324
+ "loss": 2.6847219467163086,
325
+ "step": 900
326
+ },
327
+ {
328
+ "epoch": 0.06201550387596899,
329
+ "grad_norm": 0.435546875,
330
+ "learning_rate": 0.0007,
331
+ "loss": 2.6599369049072266,
332
+ "step": 920
333
+ },
334
+ {
335
+ "epoch": 0.06336366700370745,
336
+ "grad_norm": 0.455078125,
337
+ "learning_rate": 0.0007,
338
+ "loss": 2.6465431213378907,
339
+ "step": 940
340
+ },
341
+ {
342
+ "epoch": 0.06471183013144591,
343
+ "grad_norm": 0.455078125,
344
+ "learning_rate": 0.0007,
345
+ "loss": 2.623439979553223,
346
+ "step": 960
347
+ },
348
+ {
349
+ "epoch": 0.06605999325918437,
350
+ "grad_norm": 0.462890625,
351
+ "learning_rate": 0.0007,
352
+ "loss": 2.609505271911621,
353
+ "step": 980
354
+ },
355
+ {
356
+ "epoch": 0.06740815638692282,
357
+ "grad_norm": 0.48828125,
358
+ "learning_rate": 0.0007,
359
+ "loss": 2.610812187194824,
360
+ "step": 1000
361
+ },
362
+ {
363
+ "epoch": 0.06875631951466127,
364
+ "grad_norm": 0.423828125,
365
+ "learning_rate": 0.0007,
366
+ "loss": 2.581887054443359,
367
+ "step": 1020
368
+ },
369
+ {
370
+ "epoch": 0.07010448264239973,
371
+ "grad_norm": 0.4453125,
372
+ "learning_rate": 0.0007,
373
+ "loss": 2.5679080963134764,
374
+ "step": 1040
375
+ },
376
+ {
377
+ "epoch": 0.07145264577013818,
378
+ "grad_norm": 0.447265625,
379
+ "learning_rate": 0.0007,
380
+ "loss": 2.55521240234375,
381
+ "step": 1060
382
+ },
383
+ {
384
+ "epoch": 0.07280080889787664,
385
+ "grad_norm": 0.4453125,
386
+ "learning_rate": 0.0007,
387
+ "loss": 2.536572074890137,
388
+ "step": 1080
389
+ },
390
+ {
391
+ "epoch": 0.0741489720256151,
392
+ "grad_norm": 0.431640625,
393
+ "learning_rate": 0.0007,
394
+ "loss": 2.5341920852661133,
395
+ "step": 1100
396
+ },
397
+ {
398
+ "epoch": 0.07549713515335356,
399
+ "grad_norm": 0.427734375,
400
+ "learning_rate": 0.0007,
401
+ "loss": 2.4994813919067385,
402
+ "step": 1120
403
+ },
404
+ {
405
+ "epoch": 0.07684529828109202,
406
+ "grad_norm": 0.421875,
407
+ "learning_rate": 0.0007,
408
+ "loss": 2.4886734008789064,
409
+ "step": 1140
410
+ },
411
+ {
412
+ "epoch": 0.07819346140883048,
413
+ "grad_norm": 0.419921875,
414
+ "learning_rate": 0.0007,
415
+ "loss": 2.4889175415039064,
416
+ "step": 1160
417
+ },
418
+ {
419
+ "epoch": 0.07954162453656892,
420
+ "grad_norm": 0.4140625,
421
+ "learning_rate": 0.0007,
422
+ "loss": 2.484502410888672,
423
+ "step": 1180
424
+ },
425
+ {
426
+ "epoch": 0.08088978766430738,
427
+ "grad_norm": 0.51171875,
428
+ "learning_rate": 0.0007,
429
+ "loss": 2.476539421081543,
430
+ "step": 1200
431
+ },
432
+ {
433
+ "epoch": 0.08223795079204584,
434
+ "grad_norm": 0.41796875,
435
+ "learning_rate": 0.0007,
436
+ "loss": 2.463785171508789,
437
+ "step": 1220
438
+ },
439
+ {
440
+ "epoch": 0.0835861139197843,
441
+ "grad_norm": 0.390625,
442
+ "learning_rate": 0.0007,
443
+ "loss": 2.4509374618530275,
444
+ "step": 1240
445
+ },
446
+ {
447
+ "epoch": 0.08493427704752275,
448
+ "grad_norm": 0.4453125,
449
+ "learning_rate": 0.0007,
450
+ "loss": 2.4406501770019533,
451
+ "step": 1260
452
+ },
453
+ {
454
+ "epoch": 0.08628244017526121,
455
+ "grad_norm": 0.41796875,
456
+ "learning_rate": 0.0007,
457
+ "loss": 2.423234748840332,
458
+ "step": 1280
459
+ },
460
+ {
461
+ "epoch": 0.08763060330299967,
462
+ "grad_norm": 0.4453125,
463
+ "learning_rate": 0.0007,
464
+ "loss": 2.415129280090332,
465
+ "step": 1300
466
+ },
467
+ {
468
+ "epoch": 0.08897876643073811,
469
+ "grad_norm": 0.41015625,
470
+ "learning_rate": 0.0007,
471
+ "loss": 2.4075996398925783,
472
+ "step": 1320
473
+ },
474
+ {
475
+ "epoch": 0.09032692955847657,
476
+ "grad_norm": 0.412109375,
477
+ "learning_rate": 0.0007,
478
+ "loss": 2.396971321105957,
479
+ "step": 1340
480
+ },
481
+ {
482
+ "epoch": 0.09167509268621503,
483
+ "grad_norm": 0.431640625,
484
+ "learning_rate": 0.0007,
485
+ "loss": 2.398550605773926,
486
+ "step": 1360
487
+ },
488
+ {
489
+ "epoch": 0.09302325581395349,
490
+ "grad_norm": 0.40234375,
491
+ "learning_rate": 0.0007,
492
+ "loss": 2.3772783279418945,
493
+ "step": 1380
494
+ },
495
+ {
496
+ "epoch": 0.09437141894169195,
497
+ "grad_norm": 0.40625,
498
+ "learning_rate": 0.0007,
499
+ "loss": 2.3700124740600588,
500
+ "step": 1400
501
+ },
502
+ {
503
+ "epoch": 0.0957195820694304,
504
+ "grad_norm": 0.421875,
505
+ "learning_rate": 0.0007,
506
+ "loss": 2.3814950942993165,
507
+ "step": 1420
508
+ },
509
+ {
510
+ "epoch": 0.09706774519716886,
511
+ "grad_norm": 0.431640625,
512
+ "learning_rate": 0.0007,
513
+ "loss": 2.3590816497802733,
514
+ "step": 1440
515
+ },
516
+ {
517
+ "epoch": 0.09841590832490732,
518
+ "grad_norm": 0.470703125,
519
+ "learning_rate": 0.0007,
520
+ "loss": 2.347888374328613,
521
+ "step": 1460
522
+ },
523
+ {
524
+ "epoch": 0.09976407145264576,
525
+ "grad_norm": 0.400390625,
526
+ "learning_rate": 0.0007,
527
+ "loss": 2.3355857849121096,
528
+ "step": 1480
529
+ },
530
+ {
531
+ "epoch": 0.10111223458038422,
532
+ "grad_norm": 0.4140625,
533
+ "learning_rate": 0.0007,
534
+ "loss": 2.3234769821166994,
535
+ "step": 1500
536
+ },
537
+ {
538
+ "epoch": 0.10246039770812268,
539
+ "grad_norm": 0.4375,
540
+ "learning_rate": 0.0007,
541
+ "loss": 2.327121543884277,
542
+ "step": 1520
543
+ },
544
+ {
545
+ "epoch": 0.10380856083586114,
546
+ "grad_norm": 0.408203125,
547
+ "learning_rate": 0.0007,
548
+ "loss": 2.3147512435913087,
549
+ "step": 1540
550
+ },
551
+ {
552
+ "epoch": 0.1051567239635996,
553
+ "grad_norm": 0.408203125,
554
+ "learning_rate": 0.0007,
555
+ "loss": 2.3023752212524413,
556
+ "step": 1560
557
+ },
558
+ {
559
+ "epoch": 0.10650488709133805,
560
+ "grad_norm": 0.408203125,
561
+ "learning_rate": 0.0007,
562
+ "loss": 2.3237445831298826,
563
+ "step": 1580
564
+ },
565
+ {
566
+ "epoch": 0.10785305021907651,
567
+ "grad_norm": 0.400390625,
568
+ "learning_rate": 0.0007,
569
+ "loss": 2.288845252990723,
570
+ "step": 1600
571
+ },
572
+ {
573
+ "epoch": 0.10920121334681497,
574
+ "grad_norm": 0.4140625,
575
+ "learning_rate": 0.0007,
576
+ "loss": 2.28785400390625,
577
+ "step": 1620
578
+ },
579
+ {
580
+ "epoch": 0.11054937647455342,
581
+ "grad_norm": 0.416015625,
582
+ "learning_rate": 0.0007,
583
+ "loss": 2.300959587097168,
584
+ "step": 1640
585
+ },
586
+ {
587
+ "epoch": 0.11189753960229187,
588
+ "grad_norm": 0.380859375,
589
+ "learning_rate": 0.0007,
590
+ "loss": 2.284528350830078,
591
+ "step": 1660
592
+ },
593
+ {
594
+ "epoch": 0.11324570273003033,
595
+ "grad_norm": 0.408203125,
596
+ "learning_rate": 0.0007,
597
+ "loss": 2.267934799194336,
598
+ "step": 1680
599
+ },
600
+ {
601
+ "epoch": 0.11459386585776879,
602
+ "grad_norm": 0.423828125,
603
+ "learning_rate": 0.0007,
604
+ "loss": 2.2802879333496096,
605
+ "step": 1700
606
+ },
607
+ {
608
+ "epoch": 0.11594202898550725,
609
+ "grad_norm": 0.4375,
610
+ "learning_rate": 0.0007,
611
+ "loss": 2.258160400390625,
612
+ "step": 1720
613
+ },
614
+ {
615
+ "epoch": 0.1172901921132457,
616
+ "grad_norm": 0.39453125,
617
+ "learning_rate": 0.0007,
618
+ "loss": 2.246730995178223,
619
+ "step": 1740
620
+ },
621
+ {
622
+ "epoch": 0.11863835524098416,
623
+ "grad_norm": 0.419921875,
624
+ "learning_rate": 0.0007,
625
+ "loss": 2.252003479003906,
626
+ "step": 1760
627
+ },
628
+ {
629
+ "epoch": 0.11998651836872262,
630
+ "grad_norm": 0.38671875,
631
+ "learning_rate": 0.0007,
632
+ "loss": 2.265170669555664,
633
+ "step": 1780
634
+ },
635
+ {
636
+ "epoch": 0.12133468149646107,
637
+ "grad_norm": 0.408203125,
638
+ "learning_rate": 0.0007,
639
+ "loss": 2.2610910415649412,
640
+ "step": 1800
641
+ },
642
+ {
643
+ "epoch": 0.12268284462419952,
644
+ "grad_norm": 0.41015625,
645
+ "learning_rate": 0.0007,
646
+ "loss": 2.2539751052856447,
647
+ "step": 1820
648
+ },
649
+ {
650
+ "epoch": 0.12403100775193798,
651
+ "grad_norm": 0.404296875,
652
+ "learning_rate": 0.0007,
653
+ "loss": 2.220688819885254,
654
+ "step": 1840
655
+ },
656
+ {
657
+ "epoch": 0.12537917087967643,
658
+ "grad_norm": 0.376953125,
659
+ "learning_rate": 0.0007,
660
+ "loss": 2.222416305541992,
661
+ "step": 1860
662
+ },
663
+ {
664
+ "epoch": 0.1267273340074149,
665
+ "grad_norm": 0.3828125,
666
+ "learning_rate": 0.0007,
667
+ "loss": 2.2135461807250976,
668
+ "step": 1880
669
+ },
670
+ {
671
+ "epoch": 0.12807549713515334,
672
+ "grad_norm": 0.39453125,
673
+ "learning_rate": 0.0007,
674
+ "loss": 2.2209815979003906,
675
+ "step": 1900
676
+ },
677
+ {
678
+ "epoch": 0.12942366026289182,
679
+ "grad_norm": 0.39453125,
680
+ "learning_rate": 0.0007,
681
+ "loss": 2.199201965332031,
682
+ "step": 1920
683
+ },
684
+ {
685
+ "epoch": 0.13077182339063026,
686
+ "grad_norm": 0.41015625,
687
+ "learning_rate": 0.0007,
688
+ "loss": 2.2096010208129884,
689
+ "step": 1940
690
+ },
691
+ {
692
+ "epoch": 0.13211998651836873,
693
+ "grad_norm": 0.396484375,
694
+ "learning_rate": 0.0007,
695
+ "loss": 2.1849014282226564,
696
+ "step": 1960
697
+ },
698
+ {
699
+ "epoch": 0.13346814964610718,
700
+ "grad_norm": 0.421875,
701
+ "learning_rate": 0.0007,
702
+ "loss": 2.2073038101196287,
703
+ "step": 1980
704
+ },
705
+ {
706
+ "epoch": 0.13481631277384565,
707
+ "grad_norm": 0.373046875,
708
+ "learning_rate": 0.0007,
709
+ "loss": 2.1818737030029296,
710
+ "step": 2000
711
+ },
712
+ {
713
+ "epoch": 0.1361644759015841,
714
+ "grad_norm": 0.384765625,
715
+ "learning_rate": 0.0007,
716
+ "loss": 2.1887689590454102,
717
+ "step": 2020
718
+ },
719
+ {
720
+ "epoch": 0.13751263902932254,
721
+ "grad_norm": 0.4296875,
722
+ "learning_rate": 0.0007,
723
+ "loss": 2.1714288711547853,
724
+ "step": 2040
725
+ },
726
+ {
727
+ "epoch": 0.138860802157061,
728
+ "grad_norm": 0.392578125,
729
+ "learning_rate": 0.0007,
730
+ "loss": 2.1905704498291017,
731
+ "step": 2060
732
+ },
733
+ {
734
+ "epoch": 0.14020896528479945,
735
+ "grad_norm": 0.388671875,
736
+ "learning_rate": 0.0007,
737
+ "loss": 2.168014717102051,
738
+ "step": 2080
739
+ },
740
+ {
741
+ "epoch": 0.14155712841253792,
742
+ "grad_norm": 0.37890625,
743
+ "learning_rate": 0.0007,
744
+ "loss": 2.157928466796875,
745
+ "step": 2100
746
+ },
747
+ {
748
+ "epoch": 0.14290529154027637,
749
+ "grad_norm": 0.376953125,
750
+ "learning_rate": 0.0007,
751
+ "loss": 2.1686134338378906,
752
+ "step": 2120
753
+ },
754
+ {
755
+ "epoch": 0.14425345466801484,
756
+ "grad_norm": 0.376953125,
757
+ "learning_rate": 0.0007,
758
+ "loss": 2.167608070373535,
759
+ "step": 2140
760
+ },
761
+ {
762
+ "epoch": 0.14560161779575329,
763
+ "grad_norm": 0.3828125,
764
+ "learning_rate": 0.0007,
765
+ "loss": 2.152691650390625,
766
+ "step": 2160
767
+ },
768
+ {
769
+ "epoch": 0.14694978092349173,
770
+ "grad_norm": 0.388671875,
771
+ "learning_rate": 0.0007,
772
+ "loss": 2.1620912551879883,
773
+ "step": 2180
774
+ },
775
+ {
776
+ "epoch": 0.1482979440512302,
777
+ "grad_norm": 0.388671875,
778
+ "learning_rate": 0.0007,
779
+ "loss": 2.1393999099731444,
780
+ "step": 2200
781
+ },
782
+ {
783
+ "epoch": 0.14964610717896865,
784
+ "grad_norm": 0.375,
785
+ "learning_rate": 0.0007,
786
+ "loss": 2.14599494934082,
787
+ "step": 2220
788
+ },
789
+ {
790
+ "epoch": 0.15099427030670712,
791
+ "grad_norm": 0.3984375,
792
+ "learning_rate": 0.0007,
793
+ "loss": 2.1544084548950195,
794
+ "step": 2240
795
+ },
796
+ {
797
+ "epoch": 0.15234243343444556,
798
+ "grad_norm": 0.3984375,
799
+ "learning_rate": 0.0007,
800
+ "loss": 2.132464599609375,
801
+ "step": 2260
802
+ },
803
+ {
804
+ "epoch": 0.15369059656218403,
805
+ "grad_norm": 0.373046875,
806
+ "learning_rate": 0.0007,
807
+ "loss": 2.143385314941406,
808
+ "step": 2280
809
+ },
810
+ {
811
+ "epoch": 0.15503875968992248,
812
+ "grad_norm": 0.376953125,
813
+ "learning_rate": 0.0007,
814
+ "loss": 2.1241661071777345,
815
+ "step": 2300
816
+ },
817
+ {
818
+ "epoch": 0.15638692281766095,
819
+ "grad_norm": 0.384765625,
820
+ "learning_rate": 0.0007,
821
+ "loss": 2.135605812072754,
822
+ "step": 2320
823
+ },
824
+ {
825
+ "epoch": 0.1577350859453994,
826
+ "grad_norm": 0.37109375,
827
+ "learning_rate": 0.0007,
828
+ "loss": 2.125710105895996,
829
+ "step": 2340
830
+ },
831
+ {
832
+ "epoch": 0.15908324907313784,
833
+ "grad_norm": 0.3671875,
834
+ "learning_rate": 0.0007,
835
+ "loss": 2.1266489028930664,
836
+ "step": 2360
837
+ },
838
+ {
839
+ "epoch": 0.1604314122008763,
840
+ "grad_norm": 0.376953125,
841
+ "learning_rate": 0.0007,
842
+ "loss": 2.1287538528442385,
843
+ "step": 2380
844
+ },
845
+ {
846
+ "epoch": 0.16177957532861476,
847
+ "grad_norm": 0.384765625,
848
+ "learning_rate": 0.0007,
849
+ "loss": 2.1111021041870117,
850
+ "step": 2400
851
+ }
852
+ ],
853
+ "logging_steps": 20,
854
+ "max_steps": 2500,
855
+ "num_input_tokens_seen": 0,
856
+ "num_train_epochs": 1,
857
+ "save_steps": 100,
858
+ "stateful_callbacks": {
859
+ "TrainerControl": {
860
+ "args": {
861
+ "should_epoch_stop": false,
862
+ "should_evaluate": false,
863
+ "should_log": false,
864
+ "should_save": true,
865
+ "should_training_stop": false
866
+ },
867
+ "attributes": {}
868
+ }
869
+ },
870
+ "total_flos": 7743081126297600.0,
871
+ "train_batch_size": 64,
872
+ "trial_name": null,
873
+ "trial_params": null
874
+ }
zain/Activation/out/glu-waleed-100L_run/checkpoint-2400/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c75eb6d8bf9c953837c430a69915c00b95cc26c05d95e0643e1a66470a65c0cf
3
+ size 4920
zain/Activation/out/glu-waleed-100L_run/checkpoint-2500/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "waleed",
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-waleed-100L_run/checkpoint-2500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:27bb0a44b914edc6e9abc1eb3ed7c6a31b6317edf4cb49574f483874c9d849e9
3
+ size 33967272
zain/Activation/out/glu-waleed-100L_run/checkpoint-2500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0d99fff25475cd7ff67ac8471ba32d115108da4608fd582b0536b51f0b0851b7
3
+ size 68504996
zain/Activation/out/glu-waleed-100L_run/checkpoint-2500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3cf9097d4513154245c48236b6ec5137b7ee2a21c9f58f2cba798ea275c6026f
3
+ size 14244
zain/Activation/out/glu-waleed-100L_run/checkpoint-2500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fae3f7a9f86d9663c06f6e22221bc9e6879cd9691e2b1bba04ee6412e2572f45
3
+ size 1064