w-ahmad commited on
Commit
e9643c6
·
verified ·
1 Parent(s): 2d248ba

Auto upload zain 2026-08-12T23:05:39.337686 (part 2)

Browse files
Files changed (47) hide show
  1. zain/Activation/out/glu-waleed-100L_run/checkpoint-1400/training_args.bin +3 -0
  2. zain/Activation/out/glu-waleed-100L_run/checkpoint-1500/config.json +36 -0
  3. zain/Activation/out/glu-waleed-100L_run/checkpoint-1500/model.safetensors +3 -0
  4. zain/Activation/out/glu-waleed-100L_run/checkpoint-1500/optimizer.pt +3 -0
  5. zain/Activation/out/glu-waleed-100L_run/checkpoint-1500/rng_state.pth +3 -0
  6. zain/Activation/out/glu-waleed-100L_run/checkpoint-1500/scheduler.pt +3 -0
  7. zain/Activation/out/glu-waleed-100L_run/checkpoint-1500/tokenizer.json +0 -0
  8. zain/Activation/out/glu-waleed-100L_run/checkpoint-1500/tokenizer_config.json +13 -0
  9. zain/Activation/out/glu-waleed-100L_run/checkpoint-1500/trainer_state.json +559 -0
  10. zain/Activation/out/glu-waleed-100L_run/checkpoint-1500/training_args.bin +3 -0
  11. zain/Activation/out/glu-waleed-100L_run/checkpoint-1600/config.json +36 -0
  12. zain/Activation/out/glu-waleed-100L_run/checkpoint-1600/model.safetensors +3 -0
  13. zain/Activation/out/glu-waleed-100L_run/checkpoint-1600/optimizer.pt +3 -0
  14. zain/Activation/out/glu-waleed-100L_run/checkpoint-1600/rng_state.pth +3 -0
  15. zain/Activation/out/glu-waleed-100L_run/checkpoint-1600/scheduler.pt +3 -0
  16. zain/Activation/out/glu-waleed-100L_run/checkpoint-1600/tokenizer.json +0 -0
  17. zain/Activation/out/glu-waleed-100L_run/checkpoint-1600/tokenizer_config.json +13 -0
  18. zain/Activation/out/glu-waleed-100L_run/checkpoint-1600/trainer_state.json +594 -0
  19. zain/Activation/out/glu-waleed-100L_run/checkpoint-1600/training_args.bin +3 -0
  20. zain/Activation/out/glu-waleed-100L_run/checkpoint-1700/config.json +36 -0
  21. zain/Activation/out/glu-waleed-100L_run/checkpoint-1700/model.safetensors +3 -0
  22. zain/Activation/out/glu-waleed-100L_run/checkpoint-1700/optimizer.pt +3 -0
  23. zain/Activation/out/glu-waleed-100L_run/checkpoint-1700/rng_state.pth +3 -0
  24. zain/Activation/out/glu-waleed-100L_run/checkpoint-1700/scheduler.pt +3 -0
  25. zain/Activation/out/glu-waleed-100L_run/checkpoint-1700/tokenizer.json +0 -0
  26. zain/Activation/out/glu-waleed-100L_run/checkpoint-1700/tokenizer_config.json +13 -0
  27. zain/Activation/out/glu-waleed-100L_run/checkpoint-1700/trainer_state.json +629 -0
  28. zain/Activation/out/glu-waleed-100L_run/checkpoint-1700/training_args.bin +3 -0
  29. zain/Activation/out/glu-waleed-100L_run/checkpoint-1800/config.json +36 -0
  30. zain/Activation/out/glu-waleed-100L_run/checkpoint-1800/model.safetensors +3 -0
  31. zain/Activation/out/glu-waleed-100L_run/checkpoint-1800/optimizer.pt +3 -0
  32. zain/Activation/out/glu-waleed-100L_run/checkpoint-1800/rng_state.pth +3 -0
  33. zain/Activation/out/glu-waleed-100L_run/checkpoint-1800/scheduler.pt +3 -0
  34. zain/Activation/out/glu-waleed-100L_run/checkpoint-1800/tokenizer.json +0 -0
  35. zain/Activation/out/glu-waleed-100L_run/checkpoint-1800/tokenizer_config.json +13 -0
  36. zain/Activation/out/glu-waleed-100L_run/checkpoint-1800/trainer_state.json +664 -0
  37. zain/Activation/out/glu-waleed-100L_run/checkpoint-1800/training_args.bin +3 -0
  38. zain/Activation/out/glu-waleed-100L_run/checkpoint-1900/config.json +36 -0
  39. zain/Activation/out/glu-waleed-100L_run/checkpoint-1900/model.safetensors +3 -0
  40. zain/Activation/out/glu-waleed-100L_run/checkpoint-1900/optimizer.pt +3 -0
  41. zain/Activation/out/glu-waleed-100L_run/checkpoint-1900/rng_state.pth +3 -0
  42. zain/Activation/out/glu-waleed-100L_run/checkpoint-1900/scheduler.pt +3 -0
  43. zain/Activation/out/glu-waleed-100L_run/checkpoint-1900/tokenizer.json +0 -0
  44. zain/Activation/out/glu-waleed-100L_run/checkpoint-1900/tokenizer_config.json +13 -0
  45. zain/Activation/out/glu-waleed-100L_run/checkpoint-1900/trainer_state.json +699 -0
  46. zain/Activation/out/glu-waleed-100L_run/checkpoint-1900/training_args.bin +3 -0
  47. zain/Activation/out/glu-waleed-100L_run/training_log.jsonl +50 -0
zain/Activation/out/glu-waleed-100L_run/checkpoint-1400/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-1500/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-1500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9d125cad2a869e1da5da274dcfc269076a7b158bf36aff5a5387ea4fe8e3ffe9
3
+ size 33967272
zain/Activation/out/glu-waleed-100L_run/checkpoint-1500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2064ddcea155b7df3a04257359a02785041b639ead83be21b131863ce302a6ca
3
+ size 68504996
zain/Activation/out/glu-waleed-100L_run/checkpoint-1500/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-1500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a37538486c6fa3324bc6a6a3b2ad67db8646e004089a61bb15e3c6cba879c55
3
+ size 1064
zain/Activation/out/glu-waleed-100L_run/checkpoint-1500/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-waleed-100L_run/checkpoint-1500/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-1500/trainer_state.json ADDED
@@ -0,0 +1,559 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.10111223458038422,
6
+ "eval_steps": 2498,
7
+ "global_step": 1500,
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
+ "logging_steps": 20,
539
+ "max_steps": 2500,
540
+ "num_input_tokens_seen": 0,
541
+ "num_train_epochs": 1,
542
+ "save_steps": 100,
543
+ "stateful_callbacks": {
544
+ "TrainerControl": {
545
+ "args": {
546
+ "should_epoch_stop": false,
547
+ "should_evaluate": false,
548
+ "should_log": false,
549
+ "should_save": true,
550
+ "should_training_stop": false
551
+ },
552
+ "attributes": {}
553
+ }
554
+ },
555
+ "total_flos": 4839425703936000.0,
556
+ "train_batch_size": 64,
557
+ "trial_name": null,
558
+ "trial_params": null
559
+ }
zain/Activation/out/glu-waleed-100L_run/checkpoint-1500/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-1600/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-1600/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:edda8a4d56f4b575eb048fe6e29804823cddfecbee19016606d8da6aacb4e85b
3
+ size 33967272
zain/Activation/out/glu-waleed-100L_run/checkpoint-1600/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6651f82a71addb8202460ba2785d8c88151ca52cdcd0589cdaa8539dd228bece
3
+ size 68504996
zain/Activation/out/glu-waleed-100L_run/checkpoint-1600/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-1600/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a5207460c1b01907ad3d1f40124c12e85beccaebe656a961bcd0bff18a30ffd6
3
+ size 1064
zain/Activation/out/glu-waleed-100L_run/checkpoint-1600/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-waleed-100L_run/checkpoint-1600/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-1600/trainer_state.json ADDED
@@ -0,0 +1,594 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.10785305021907651,
6
+ "eval_steps": 2498,
7
+ "global_step": 1600,
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
+ "logging_steps": 20,
574
+ "max_steps": 2500,
575
+ "num_input_tokens_seen": 0,
576
+ "num_train_epochs": 1,
577
+ "save_steps": 100,
578
+ "stateful_callbacks": {
579
+ "TrainerControl": {
580
+ "args": {
581
+ "should_epoch_stop": false,
582
+ "should_evaluate": false,
583
+ "should_log": false,
584
+ "should_save": true,
585
+ "should_training_stop": false
586
+ },
587
+ "attributes": {}
588
+ }
589
+ },
590
+ "total_flos": 5162054084198400.0,
591
+ "train_batch_size": 64,
592
+ "trial_name": null,
593
+ "trial_params": null
594
+ }
zain/Activation/out/glu-waleed-100L_run/checkpoint-1600/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-1700/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-1700/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:006cfded52fc110c49550781ff583cc5de3744320977b3c937d81fb25b81c4b0
3
+ size 33967272
zain/Activation/out/glu-waleed-100L_run/checkpoint-1700/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b9a4d02c124748b600a37d244282051c4a3c74241d904c47a90d011aa56d117f
3
+ size 68504996
zain/Activation/out/glu-waleed-100L_run/checkpoint-1700/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-1700/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:da239a902a87900de7861a410c642a5ae70d481b0020a3040c363cc3ced9b7a4
3
+ size 1064
zain/Activation/out/glu-waleed-100L_run/checkpoint-1700/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-waleed-100L_run/checkpoint-1700/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-1700/trainer_state.json ADDED
@@ -0,0 +1,629 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.11459386585776879,
6
+ "eval_steps": 2498,
7
+ "global_step": 1700,
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
+ "logging_steps": 20,
609
+ "max_steps": 2500,
610
+ "num_input_tokens_seen": 0,
611
+ "num_train_epochs": 1,
612
+ "save_steps": 100,
613
+ "stateful_callbacks": {
614
+ "TrainerControl": {
615
+ "args": {
616
+ "should_epoch_stop": false,
617
+ "should_evaluate": false,
618
+ "should_log": false,
619
+ "should_save": true,
620
+ "should_training_stop": false
621
+ },
622
+ "attributes": {}
623
+ }
624
+ },
625
+ "total_flos": 5484682464460800.0,
626
+ "train_batch_size": 64,
627
+ "trial_name": null,
628
+ "trial_params": null
629
+ }
zain/Activation/out/glu-waleed-100L_run/checkpoint-1700/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-1800/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-1800/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d587a88a6dc402b66467c2d09de836281143e53d8092b04e68059c7863097d7b
3
+ size 33967272
zain/Activation/out/glu-waleed-100L_run/checkpoint-1800/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:55043b24e87cebc1830c015e71b35e699cb9d7d1240104095d608e472861630d
3
+ size 68504996
zain/Activation/out/glu-waleed-100L_run/checkpoint-1800/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-1800/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8214bf4c471bc465884f1c219edb27abec42cbdf74a9a2fe9320000a4ce60c16
3
+ size 1064
zain/Activation/out/glu-waleed-100L_run/checkpoint-1800/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-waleed-100L_run/checkpoint-1800/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-1800/trainer_state.json ADDED
@@ -0,0 +1,664 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.12133468149646107,
6
+ "eval_steps": 2498,
7
+ "global_step": 1800,
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
+ "logging_steps": 20,
644
+ "max_steps": 2500,
645
+ "num_input_tokens_seen": 0,
646
+ "num_train_epochs": 1,
647
+ "save_steps": 100,
648
+ "stateful_callbacks": {
649
+ "TrainerControl": {
650
+ "args": {
651
+ "should_epoch_stop": false,
652
+ "should_evaluate": false,
653
+ "should_log": false,
654
+ "should_save": true,
655
+ "should_training_stop": false
656
+ },
657
+ "attributes": {}
658
+ }
659
+ },
660
+ "total_flos": 5807310844723200.0,
661
+ "train_batch_size": 64,
662
+ "trial_name": null,
663
+ "trial_params": null
664
+ }
zain/Activation/out/glu-waleed-100L_run/checkpoint-1800/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-1900/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-1900/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef09fd2b3827ce07b8e27ba9208e4606f6fa42a043afacb3066dac67771fb142
3
+ size 33967272
zain/Activation/out/glu-waleed-100L_run/checkpoint-1900/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a814ac18104474835980ac4b2911fc6210880e0060de447622968b55e0772fa8
3
+ size 68504996
zain/Activation/out/glu-waleed-100L_run/checkpoint-1900/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-1900/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa8588bc4a247e7962c4619ae7e7ba14fb542cc23b88b7c743a799982e027413
3
+ size 1064
zain/Activation/out/glu-waleed-100L_run/checkpoint-1900/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-waleed-100L_run/checkpoint-1900/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-1900/trainer_state.json ADDED
@@ -0,0 +1,699 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.12807549713515334,
6
+ "eval_steps": 2498,
7
+ "global_step": 1900,
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
+ "logging_steps": 20,
679
+ "max_steps": 2500,
680
+ "num_input_tokens_seen": 0,
681
+ "num_train_epochs": 1,
682
+ "save_steps": 100,
683
+ "stateful_callbacks": {
684
+ "TrainerControl": {
685
+ "args": {
686
+ "should_epoch_stop": false,
687
+ "should_evaluate": false,
688
+ "should_log": false,
689
+ "should_save": true,
690
+ "should_training_stop": false
691
+ },
692
+ "attributes": {}
693
+ }
694
+ },
695
+ "total_flos": 6129939224985600.0,
696
+ "train_batch_size": 64,
697
+ "trial_name": null,
698
+ "trial_params": null
699
+ }
zain/Activation/out/glu-waleed-100L_run/checkpoint-1900/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/training_log.jsonl CHANGED
@@ -47,3 +47,53 @@
47
  {"step": 940, "epoch": 0.06336366700370745, "timestamp": 1786575619.2992687, "loss": 2.6465431213378907, "grad_norm": 0.455078125, "learning_rate": 0.0007, "train/total_time_seconds": 243.2623048685491, "train/time_per_step_avg": 0.261282008998096, "train/epoch_time_elapsed": 292.9475435130298, "train/estimated_remaining_minutes": 6.7285318367896565}
48
  {"step": 960, "epoch": 0.06471183013144591, "timestamp": 1786575625.4047773, "loss": 2.623439979553223, "grad_norm": 0.455078125, "learning_rate": 0.0007, "train/total_time_seconds": 248.3307976089418, "train/time_per_step_avg": 0.2583442800119519, "train/epoch_time_elapsed": 299.0530513934791, "train/estimated_remaining_minutes": 6.639399797183513}
49
  {"step": 980, "epoch": 0.06605999325918437, "timestamp": 1786575631.5448115, "loss": 2.609505271911621, "grad_norm": 0.462890625, "learning_rate": 0.0007, "train/total_time_seconds": 253.4427866525948, "train/time_per_step_avg": 0.2579416869953275, "train/epoch_time_elapsed": 305.19308672472835, "train/estimated_remaining_minutes": 6.55158223999905}
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
47
  {"step": 940, "epoch": 0.06336366700370745, "timestamp": 1786575619.2992687, "loss": 2.6465431213378907, "grad_norm": 0.455078125, "learning_rate": 0.0007, "train/total_time_seconds": 243.2623048685491, "train/time_per_step_avg": 0.261282008998096, "train/epoch_time_elapsed": 292.9475435130298, "train/estimated_remaining_minutes": 6.7285318367896565}
48
  {"step": 960, "epoch": 0.06471183013144591, "timestamp": 1786575625.4047773, "loss": 2.623439979553223, "grad_norm": 0.455078125, "learning_rate": 0.0007, "train/total_time_seconds": 248.3307976089418, "train/time_per_step_avg": 0.2583442800119519, "train/epoch_time_elapsed": 299.0530513934791, "train/estimated_remaining_minutes": 6.639399797183513}
49
  {"step": 980, "epoch": 0.06605999325918437, "timestamp": 1786575631.5448115, "loss": 2.609505271911621, "grad_norm": 0.462890625, "learning_rate": 0.0007, "train/total_time_seconds": 253.4427866525948, "train/time_per_step_avg": 0.2579416869953275, "train/epoch_time_elapsed": 305.19308672472835, "train/estimated_remaining_minutes": 6.55158223999905}
50
+ {"step": 1000, "epoch": 0.06740815638692282, "timestamp": 1786575637.7622666, "loss": 2.610812187194824, "grad_norm": 0.48828125, "learning_rate": 0.0007, "train/total_time_seconds": 258.6546031832695, "train/time_per_step_avg": 0.2582899022474885, "train/epoch_time_elapsed": 311.4105418436229, "train/estimated_remaining_minutes": 6.466365079581737}
51
+ {"step": 1020, "epoch": 0.06875631951466127, "timestamp": 1786575644.1526191, "loss": 2.581887054443359, "grad_norm": 0.423828125, "learning_rate": 0.0007, "train/total_time_seconds": 263.77820168435574, "train/time_per_step_avg": 0.25763155523687603, "train/epoch_time_elapsed": 317.80089434608817, "train/estimated_remaining_minutes": 6.378949975373308}
52
+ {"step": 1040, "epoch": 0.07010448264239973, "timestamp": 1786575650.335695, "loss": 2.5679080963134764, "grad_norm": 0.4453125, "learning_rate": 0.0007, "train/total_time_seconds": 268.9611978940666, "train/time_per_step_avg": 0.25698893025517466, "train/epoch_time_elapsed": 323.9839700423181, "train/estimated_remaining_minutes": 6.293002386623994}
53
+ {"step": 1060, "epoch": 0.07145264577013818, "timestamp": 1786575656.4630058, "loss": 2.55521240234375, "grad_norm": 0.447265625, "learning_rate": 0.0007, "train/total_time_seconds": 274.10039650276303, "train/time_per_step_avg": 0.2576959889382124, "train/epoch_time_elapsed": 330.1112811565399, "train/estimated_remaining_minutes": 6.206046713270107}
54
+ {"step": 1080, "epoch": 0.07280080889787664, "timestamp": 1786575662.6544578, "loss": 2.536572074890137, "grad_norm": 0.4453125, "learning_rate": 0.0007, "train/total_time_seconds": 279.27789648249745, "train/time_per_step_avg": 0.25835109829902647, "train/epoch_time_elapsed": 336.3027331903577, "train/estimated_remaining_minutes": 6.119978595758432}
55
+ {"step": 1100, "epoch": 0.0741489720256151, "timestamp": 1786575668.7768958, "loss": 2.5341920852661133, "grad_norm": 0.431640625, "learning_rate": 0.0007, "train/total_time_seconds": 284.409443564713, "train/time_per_step_avg": 0.257548403814435, "train/epoch_time_elapsed": 342.4251710213721, "train/estimated_remaining_minutes": 6.032927590766638}
56
+ {"step": 1120, "epoch": 0.07549713515335356, "timestamp": 1786575675.368542, "loss": 2.4994813919067385, "grad_norm": 0.427734375, "learning_rate": 0.0007, "train/total_time_seconds": 289.73538291454315, "train/time_per_step_avg": 0.25957181230187415, "train/epoch_time_elapsed": 349.0168175585568, "train/estimated_remaining_minutes": 5.949923041995083}
57
+ {"step": 1140, "epoch": 0.07684529828109202, "timestamp": 1786575681.4954743, "loss": 2.4886734008789064, "grad_norm": 0.421875, "learning_rate": 0.0007, "train/total_time_seconds": 294.882439866662, "train/time_per_step_avg": 0.25921241972595455, "train/epoch_time_elapsed": 355.1437474861741, "train/estimated_remaining_minutes": 5.86315962307983}
58
+ {"step": 1160, "epoch": 0.07819346140883048, "timestamp": 1786575687.671574, "loss": 2.4889175415039064, "grad_norm": 0.419921875, "learning_rate": 0.0007, "train/total_time_seconds": 300.0251535065472, "train/time_per_step_avg": 0.2592475700378418, "train/epoch_time_elapsed": 361.3198497183621, "train/estimated_remaining_minutes": 5.776346346246742}
59
+ {"step": 1180, "epoch": 0.07954162453656892, "timestamp": 1786575693.7874627, "loss": 2.484502410888672, "grad_norm": 0.4140625, "learning_rate": 0.0007, "train/total_time_seconds": 305.1488474570215, "train/time_per_step_avg": 0.2587095097452402, "train/epoch_time_elapsed": 367.43573766201735, "train/estimated_remaining_minutes": 5.689215800046163}
60
+ {"step": 1200, "epoch": 0.08088978766430738, "timestamp": 1786575699.9657428, "loss": 2.476539421081543, "grad_norm": 0.51171875, "learning_rate": 0.0007, "train/total_time_seconds": 310.31584874913096, "train/time_per_step_avg": 0.25906405184417963, "train/epoch_time_elapsed": 373.6140180788934, "train/estimated_remaining_minutes": 5.602925046859309}
61
+ {"step": 1220, "epoch": 0.08223795079204584, "timestamp": 1786575706.4236705, "loss": 2.463785171508789, "grad_norm": 0.41796875, "learning_rate": 0.0007, "train/total_time_seconds": 315.4617696851492, "train/time_per_step_avg": 0.2572638677060604, "train/epoch_time_elapsed": 380.07194494083524, "train/estimated_remaining_minutes": 5.516271382472554}
62
+ {"step": 1240, "epoch": 0.0835861139197843, "timestamp": 1786575712.5356638, "loss": 2.4509374618530275, "grad_norm": 0.390625, "learning_rate": 0.0007, "train/total_time_seconds": 320.5749963186681, "train/time_per_step_avg": 0.256925564520061, "train/epoch_time_elapsed": 386.18393890559673, "train/estimated_remaining_minutes": 5.429092679590348}
63
+ {"step": 1260, "epoch": 0.08493427704752275, "timestamp": 1786575718.7027848, "loss": 2.4406501770019533, "grad_norm": 0.4453125, "learning_rate": 0.0007, "train/total_time_seconds": 325.7314862795174, "train/time_per_step_avg": 0.257063327729702, "train/epoch_time_elapsed": 392.35105869174004, "train/estimated_remaining_minutes": 5.342685753791027}
64
+ {"step": 1280, "epoch": 0.08628244017526121, "timestamp": 1786575724.8584175, "loss": 2.423234748840332, "grad_norm": 0.41796875, "learning_rate": 0.0007, "train/total_time_seconds": 330.88962161540985, "train/time_per_step_avg": 0.25740774158388374, "train/epoch_time_elapsed": 398.50669237971306, "train/estimated_remaining_minutes": 5.256319510036459}
65
+ {"step": 1300, "epoch": 0.08763060330299967, "timestamp": 1786575731.0391603, "loss": 2.415129280090332, "grad_norm": 0.4453125, "learning_rate": 0.0007, "train/total_time_seconds": 336.064381621778, "train/time_per_step_avg": 0.25748532872647045, "train/epoch_time_elapsed": 404.68743501231074, "train/estimated_remaining_minutes": 5.170221255719662}
66
+ {"step": 1320, "epoch": 0.08897876643073811, "timestamp": 1786575737.5706024, "loss": 2.4075996398925783, "grad_norm": 0.41015625, "learning_rate": 0.0007, "train/total_time_seconds": 341.2818893864751, "train/time_per_step_avg": 0.25820119701325894, "train/epoch_time_elapsed": 411.2188755236566, "train/estimated_remaining_minutes": 5.08475542267728}
67
+ {"step": 1340, "epoch": 0.09032692955847657, "timestamp": 1786575743.6985369, "loss": 2.396971321105957, "grad_norm": 0.412109375, "learning_rate": 0.0007, "train/total_time_seconds": 346.4173893928528, "train/time_per_step_avg": 0.25842393074184655, "train/epoch_time_elapsed": 417.34681187570095, "train/estimated_remaining_minutes": 4.998061837011308}
68
+ {"step": 1360, "epoch": 0.09167509268621503, "timestamp": 1786575749.913696, "loss": 2.398550605773926, "grad_norm": 0.431640625, "learning_rate": 0.0007, "train/total_time_seconds": 351.63109173998237, "train/time_per_step_avg": 0.25899605460464953, "train/epoch_time_elapsed": 423.56197110563517, "train/estimated_remaining_minutes": 4.912493193426225}
69
+ {"step": 1380, "epoch": 0.09302325581395349, "timestamp": 1786575756.0410607, "loss": 2.3772783279418945, "grad_norm": 0.40234375, "learning_rate": 0.0007, "train/total_time_seconds": 356.76987397298217, "train/time_per_step_avg": 0.2588025235757232, "train/epoch_time_elapsed": 429.6893359422684, "train/estimated_remaining_minutes": 4.825872691421981}
70
+ {"step": 1400, "epoch": 0.09437141894169195, "timestamp": 1786575762.1742163, "loss": 2.3700124740600588, "grad_norm": 0.40625, "learning_rate": 0.0007, "train/total_time_seconds": 361.9081694111228, "train/time_per_step_avg": 0.25843787789344785, "train/epoch_time_elapsed": 435.8224910646677, "train/estimated_remaining_minutes": 4.739273647050418}
71
+ {"step": 1420, "epoch": 0.0957195820694304, "timestamp": 1786575768.7868671, "loss": 2.3814950942993165, "grad_norm": 0.421875, "learning_rate": 0.0007, "train/total_time_seconds": 367.26074535399675, "train/time_per_step_avg": 0.2597885596752167, "train/epoch_time_elapsed": 442.4351426139474, "train/estimated_remaining_minutes": 4.6554178988534805}
72
+ {"step": 1440, "epoch": 0.09706774519716886, "timestamp": 1786575774.8903534, "loss": 2.3590816497802733, "grad_norm": 0.431640625, "learning_rate": 0.0007, "train/total_time_seconds": 372.37719567120075, "train/time_per_step_avg": 0.2595980627834797, "train/epoch_time_elapsed": 448.53862861171365, "train/estimated_remaining_minutes": 4.568516520966122}
73
+ {"step": 1460, "epoch": 0.09841590832490732, "timestamp": 1786575781.0615234, "loss": 2.347888374328613, "grad_norm": 0.470703125, "learning_rate": 0.0007, "train/total_time_seconds": 377.5522587969899, "train/time_per_step_avg": 0.2592116705700755, "train/epoch_time_elapsed": 454.7097984999418, "train/estimated_remaining_minutes": 4.482355583891205}
74
+ {"step": 1480, "epoch": 0.09976407145264576, "timestamp": 1786575787.2440212, "loss": 2.3355857849121096, "grad_norm": 0.400390625, "learning_rate": 0.0007, "train/total_time_seconds": 382.6989061422646, "train/time_per_step_avg": 0.25929032169282434, "train/epoch_time_elapsed": 460.892296474427, "train/estimated_remaining_minutes": 4.395865813796282}
75
+ {"step": 1500, "epoch": 0.10111223458038422, "timestamp": 1786575793.4448934, "loss": 2.3234769821166994, "grad_norm": 0.4140625, "learning_rate": 0.0007, "train/total_time_seconds": 387.88587468117476, "train/time_per_step_avg": 0.25977705270051954, "train/epoch_time_elapsed": 467.09316850453615, "train/estimated_remaining_minutes": 4.309843052013052}
76
+ {"step": 1520, "epoch": 0.10246039770812268, "timestamp": 1786575799.8463821, "loss": 2.327121543884277, "grad_norm": 0.4375, "learning_rate": 0.0007, "train/total_time_seconds": 392.9994839541614, "train/time_per_step_avg": 0.2573873860016465, "train/epoch_time_elapsed": 473.49465545266867, "train/estimated_remaining_minutes": 4.223020770560067}
77
+ {"step": 1540, "epoch": 0.10380856083586114, "timestamp": 1786575805.962222, "loss": 2.3147512435913087, "grad_norm": 0.408203125, "learning_rate": 0.0007, "train/total_time_seconds": 398.12631433457136, "train/time_per_step_avg": 0.25749118663370607, "train/epoch_time_elapsed": 479.61049706861377, "train/estimated_remaining_minutes": 4.136377291787754}
78
+ {"step": 1560, "epoch": 0.1051567239635996, "timestamp": 1786575812.2014282, "loss": 2.3023752212524413, "grad_norm": 0.408203125, "learning_rate": 0.0007, "train/total_time_seconds": 403.3427828364074, "train/time_per_step_avg": 0.25790524039417506, "train/epoch_time_elapsed": 485.8497031405568, "train/estimated_remaining_minutes": 4.050664699425459}
79
+ {"step": 1580, "epoch": 0.10650488709133805, "timestamp": 1786575818.3715289, "loss": 2.3237445831298826, "grad_norm": 0.408203125, "learning_rate": 0.0007, "train/total_time_seconds": 408.52200120687485, "train/time_per_step_avg": 0.2582309506461024, "train/epoch_time_elapsed": 492.0198044255376, "train/estimated_remaining_minutes": 3.9645595053831735}
80
+ {"step": 1600, "epoch": 0.10785305021907651, "timestamp": 1786575824.484237, "loss": 2.288845252990723, "grad_norm": 0.400390625, "learning_rate": 0.0007, "train/total_time_seconds": 413.6448117978871, "train/time_per_step_avg": 0.2575893711671233, "train/epoch_time_elapsed": 498.1325124576688, "train/estimated_remaining_minutes": 3.8779201106051913}
81
+ {"step": 1620, "epoch": 0.10920121334681497, "timestamp": 1786575830.864202, "loss": 2.28785400390625, "grad_norm": 0.4140625, "learning_rate": 0.0007, "train/total_time_seconds": 418.7556971721351, "train/time_per_step_avg": 0.2575621321797371, "train/epoch_time_elapsed": 504.5124769806862, "train/estimated_remaining_minutes": 3.791203842710688}
82
+ {"step": 1640, "epoch": 0.11054937647455342, "timestamp": 1786575837.059811, "loss": 2.300959587097168, "grad_norm": 0.416015625, "learning_rate": 0.0007, "train/total_time_seconds": 423.9290645904839, "train/time_per_step_avg": 0.25802750255912543, "train/epoch_time_elapsed": 510.7080861926079, "train/estimated_remaining_minutes": 3.7050710929656114}
83
+ {"step": 1660, "epoch": 0.11189753960229187, "timestamp": 1786575843.1866276, "loss": 2.284528350830078, "grad_norm": 0.380859375, "learning_rate": 0.0007, "train/total_time_seconds": 429.05558678880334, "train/time_per_step_avg": 0.2571280395239592, "train/epoch_time_elapsed": 516.834902331233, "train/estimated_remaining_minutes": 3.618541093399546}
84
+ {"step": 1680, "epoch": 0.11324570273003033, "timestamp": 1786575849.346091, "loss": 2.267934799194336, "grad_norm": 0.408203125, "learning_rate": 0.0007, "train/total_time_seconds": 434.21137834712863, "train/time_per_step_avg": 0.25689377140253783, "train/epoch_time_elapsed": 522.9943660236895, "train/estimated_remaining_minutes": 3.5322751016333873}
85
+ {"step": 1700, "epoch": 0.11459386585776879, "timestamp": 1786575855.4579034, "loss": 2.2802879333496096, "grad_norm": 0.423828125, "learning_rate": 0.0007, "train/total_time_seconds": 439.335325691849, "train/time_per_step_avg": 0.2569051389396191, "train/epoch_time_elapsed": 529.1061787717044, "train/estimated_remaining_minutes": 3.4457672603282274}
86
+ {"step": 1720, "epoch": 0.11594202898550725, "timestamp": 1786575862.0518303, "loss": 2.258160400390625, "grad_norm": 0.4375, "learning_rate": 0.0007, "train/total_time_seconds": 444.6602842658758, "train/time_per_step_avg": 0.259045870937407, "train/epoch_time_elapsed": 535.7001053728163, "train/estimated_remaining_minutes": 3.36080447410255}
87
+ {"step": 1740, "epoch": 0.1172901921132457, "timestamp": 1786575868.2333925, "loss": 2.246730995178223, "grad_norm": 0.39453125, "learning_rate": 0.0007, "train/total_time_seconds": 449.82554940506816, "train/time_per_step_avg": 0.25896484814584253, "train/epoch_time_elapsed": 541.8816677071154, "train/estimated_remaining_minutes": 3.2745921221058603}
88
+ {"step": 1760, "epoch": 0.11863835524098416, "timestamp": 1786575874.310372, "loss": 2.252003479003906, "grad_norm": 0.419921875, "learning_rate": 0.0007, "train/total_time_seconds": 454.92364130169153, "train/time_per_step_avg": 0.25868054512888194, "train/epoch_time_elapsed": 547.9586466923356, "train/estimated_remaining_minutes": 3.187911880333823}
89
+ {"step": 1780, "epoch": 0.11998651836872262, "timestamp": 1786575880.3955638, "loss": 2.265170669555664, "grad_norm": 0.38671875, "learning_rate": 0.0007, "train/total_time_seconds": 460.0269968956709, "train/time_per_step_avg": 0.25815618548542263, "train/epoch_time_elapsed": 554.0438387393951, "train/estimated_remaining_minutes": 3.1013055970494667}
90
+ {"step": 1800, "epoch": 0.12133468149646107, "timestamp": 1786575886.5105824, "loss": 2.2610910415649412, "grad_norm": 0.408203125, "learning_rate": 0.0007, "train/total_time_seconds": 465.157982558012, "train/time_per_step_avg": 0.25822656866163013, "train/epoch_time_elapsed": 560.1588573940098, "train/estimated_remaining_minutes": 3.0149128499130406}
91
+ {"step": 1820, "epoch": 0.12268284462419952, "timestamp": 1786575892.9290805, "loss": 2.2539751052856447, "grad_norm": 0.41015625, "learning_rate": 0.0007, "train/total_time_seconds": 470.25223245471716, "train/time_per_step_avg": 0.2559194818884134, "train/epoch_time_elapsed": 566.5773558542132, "train/estimated_remaining_minutes": 2.9283106050293743}
92
+ {"step": 1840, "epoch": 0.12403100775193798, "timestamp": 1786575899.0055704, "loss": 2.220688819885254, "grad_norm": 0.404296875, "learning_rate": 0.0007, "train/total_time_seconds": 475.3463321775198, "train/time_per_step_avg": 0.2552078277245164, "train/epoch_time_elapsed": 572.6538440175354, "train/estimated_remaining_minutes": 2.8417443771482165}
93
+ {"step": 1860, "epoch": 0.12537917087967643, "timestamp": 1786575905.2542048, "loss": 2.222416305541992, "grad_norm": 0.376953125, "learning_rate": 0.0007, "train/total_time_seconds": 480.54932870715857, "train/time_per_step_avg": 0.25625687405467035, "train/epoch_time_elapsed": 578.9024794436991, "train/estimated_remaining_minutes": 2.755838444198759}
94
+ {"step": 1880, "epoch": 0.1267273340074149, "timestamp": 1786575911.4729314, "loss": 2.2135461807250976, "grad_norm": 0.3828125, "learning_rate": 0.0007, "train/total_time_seconds": 485.7390873245895, "train/time_per_step_avg": 0.257120904289186, "train/epoch_time_elapsed": 585.1212064437568, "train/estimated_remaining_minutes": 2.6698425012521763}
95
+ {"step": 1900, "epoch": 0.12807549713515334, "timestamp": 1786575917.6023524, "loss": 2.2209815979003906, "grad_norm": 0.39453125, "learning_rate": 0.0007, "train/total_time_seconds": 490.88303845748305, "train/time_per_step_avg": 0.25725055899471044, "train/epoch_time_elapsed": 591.2506272792816, "train/estimated_remaining_minutes": 2.583594939249911}
96
+ {"step": 1920, "epoch": 0.12942366026289182, "timestamp": 1786575924.005875, "loss": 2.199201965332031, "grad_norm": 0.39453125, "learning_rate": 0.0007, "train/total_time_seconds": 496.01020282134414, "train/time_per_step_avg": 0.25757970366626975, "train/epoch_time_elapsed": 597.654150467366, "train/estimated_remaining_minutes": 2.4972735905935726}
97
+ {"step": 1940, "epoch": 0.13077182339063026, "timestamp": 1786575930.2003074, "loss": 2.2096010208129884, "grad_norm": 0.41015625, "learning_rate": 0.0007, "train/total_time_seconds": 501.2104085125029, "train/time_per_step_avg": 0.2586407633498311, "train/epoch_time_elapsed": 603.848582174629, "train/estimated_remaining_minutes": 2.411321552981114}
98
+ {"step": 1960, "epoch": 0.13211998651836873, "timestamp": 1786575936.374858, "loss": 2.1849014282226564, "grad_norm": 0.396484375, "learning_rate": 0.0007, "train/total_time_seconds": 506.3590369075537, "train/time_per_step_avg": 0.2580970820039511, "train/epoch_time_elapsed": 610.023133251816, "train/estimated_remaining_minutes": 2.325118026616318}
99
+ {"step": 1980, "epoch": 0.13346814964610718, "timestamp": 1786575942.5903032, "loss": 2.2073038101196287, "grad_norm": 0.421875, "learning_rate": 0.0007, "train/total_time_seconds": 511.55915205925703, "train/time_per_step_avg": 0.2582006473466754, "train/epoch_time_elapsed": 616.2385786548257, "train/estimated_remaining_minutes": 2.239147803626378}