w-ahmad commited on
Commit
1ff694b
·
verified ·
1 Parent(s): 3e3b55e

Auto upload zain 2026-08-12T23:26:11.331277

Browse files
Files changed (33) hide show
  1. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/config.json +36 -0
  2. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/model.safetensors +3 -0
  3. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/optimizer.pt +3 -0
  4. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/rng_state.pth +3 -0
  5. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/scheduler.pt +3 -0
  6. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/tokenizer.json +0 -0
  7. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/tokenizer_config.json +13 -0
  8. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/trainer_state.json +839 -0
  9. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/training_args.bin +3 -0
  10. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/config.json +36 -0
  11. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/model.safetensors +3 -0
  12. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/optimizer.pt +3 -0
  13. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/rng_state.pth +3 -0
  14. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/scheduler.pt +3 -0
  15. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/tokenizer.json +0 -0
  16. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/tokenizer_config.json +13 -0
  17. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/trainer_state.json +874 -0
  18. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/training_args.bin +3 -0
  19. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/config.json +36 -0
  20. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/model.safetensors +3 -0
  21. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/optimizer.pt +3 -0
  22. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/rng_state.pth +3 -0
  23. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/scheduler.pt +3 -0
  24. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/tokenizer.json +0 -0
  25. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/tokenizer_config.json +13 -0
  26. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/trainer_state.json +925 -0
  27. zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/training_args.bin +3 -0
  28. zain/Activation/out/glu-situglu_low-100L_run/config.json +36 -0
  29. zain/Activation/out/glu-situglu_low-100L_run/model.safetensors +3 -0
  30. zain/Activation/out/glu-situglu_low-100L_run/tokenizer.json +0 -0
  31. zain/Activation/out/glu-situglu_low-100L_run/tokenizer_config.json +13 -0
  32. zain/Activation/out/glu-situglu_low-100L_run/training_args.bin +3 -0
  33. zain/Activation/out/glu-situglu_low-100L_run/training_log.jsonl +18 -0
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "situglu_low",
3
+ "architectures": [
4
+ "TinyLlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 2,
11
+ "head_dim": 32,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 128,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 256,
16
+ "max_position_embeddings": 512,
17
+ "mlp_bias": false,
18
+ "mlp_type": "glu",
19
+ "model_type": "tiny_llama",
20
+ "num_attention_heads": 4,
21
+ "num_hidden_layers": 100,
22
+ "num_key_value_heads": 4,
23
+ "pad_token_id": 0,
24
+ "pretraining_tp": 1,
25
+ "rms_norm_eps": 1e-06,
26
+ "rope_parameters": {
27
+ "rope_theta": 10000.0,
28
+ "rope_type": "default"
29
+ },
30
+ "tie_word_embeddings": true,
31
+ "tokenizer_name": "w-ahmad/tiny-stories-tokenizer",
32
+ "transformers_version": "5.16.0.dev0",
33
+ "use_cache": false,
34
+ "vocab_size": 4096,
35
+ "waleed_beta": 10.0
36
+ }
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e80bb345310a0a0d39ff467b5609052d4077579b84dc3d2e12d86641493096a6
3
+ size 33967272
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:922f74421cabb5144e418882b5bfa50a4bc877f042bef6c707e3de32fa377ff3
3
+ size 68504996
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9d9cd6a0487226e5bd30d1846894c82af483733ab4381b75bae9c0745e05d405
3
+ size 14244
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b9bcfc4abc1cd7f19c4260bd52451981a53bf8db4a664a25c46ab7065952dd0
3
+ size 1064
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/tokenizer_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|endoftext|>",
5
+ "eos_token": "<|endoftext|>",
6
+ "errors": "replace",
7
+ "is_local": false,
8
+ "local_files_only": false,
9
+ "model_max_length": 1024,
10
+ "pad_token": "<|endoftext|>",
11
+ "tokenizer_class": "GPT2Tokenizer",
12
+ "unk_token": "<|endoftext|>"
13
+ }
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/trainer_state.json ADDED
@@ -0,0 +1,839 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.15503875968992248,
6
+ "eval_steps": 2498,
7
+ "global_step": 2300,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0013481631277384564,
14
+ "grad_norm": 2.078125,
15
+ "learning_rate": 2.66e-05,
16
+ "loss": 8.299667358398438,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.3515625,
22
+ "learning_rate": 5.46e-05,
23
+ "loss": 8.047861480712891,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 1.2734375,
29
+ "learning_rate": 8.259999999999999e-05,
30
+ "loss": 7.838032531738281,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 1.265625,
36
+ "learning_rate": 0.0001106,
37
+ "loss": 7.572964477539062,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 1.1953125,
43
+ "learning_rate": 0.0001386,
44
+ "loss": 7.233274841308594,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 1.1328125,
50
+ "learning_rate": 0.00016659999999999998,
51
+ "loss": 6.860704803466797,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 0.94921875,
57
+ "learning_rate": 0.00019460000000000001,
58
+ "loss": 6.517856597900391,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 0.8046875,
64
+ "learning_rate": 0.0002226,
65
+ "loss": 6.239397811889648,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 0.65234375,
71
+ "learning_rate": 0.00025059999999999997,
72
+ "loss": 6.039968109130859,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 0.5625,
78
+ "learning_rate": 0.0002786,
79
+ "loss": 5.863864517211914,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.01482979440512302,
84
+ "grad_norm": 0.67578125,
85
+ "learning_rate": 0.00030659999999999997,
86
+ "loss": 5.679952239990234,
87
+ "step": 220
88
+ },
89
+ {
90
+ "epoch": 0.016177957532861477,
91
+ "grad_norm": 1.484375,
92
+ "learning_rate": 0.0003346,
93
+ "loss": 5.470568084716797,
94
+ "step": 240
95
+ },
96
+ {
97
+ "epoch": 0.01752612066059993,
98
+ "grad_norm": 1.9765625,
99
+ "learning_rate": 0.00036260000000000003,
100
+ "loss": 5.266762542724609,
101
+ "step": 260
102
+ },
103
+ {
104
+ "epoch": 0.01887428378833839,
105
+ "grad_norm": 0.6015625,
106
+ "learning_rate": 0.0003906,
107
+ "loss": 5.04127197265625,
108
+ "step": 280
109
+ },
110
+ {
111
+ "epoch": 0.020222446916076844,
112
+ "grad_norm": 1.0859375,
113
+ "learning_rate": 0.0004186,
114
+ "loss": 4.803184509277344,
115
+ "step": 300
116
+ },
117
+ {
118
+ "epoch": 0.021570610043815303,
119
+ "grad_norm": 1.15625,
120
+ "learning_rate": 0.0004466,
121
+ "loss": 4.589862823486328,
122
+ "step": 320
123
+ },
124
+ {
125
+ "epoch": 0.022918773171553757,
126
+ "grad_norm": 0.6953125,
127
+ "learning_rate": 0.00047460000000000004,
128
+ "loss": 4.416217803955078,
129
+ "step": 340
130
+ },
131
+ {
132
+ "epoch": 0.024266936299292215,
133
+ "grad_norm": 0.8046875,
134
+ "learning_rate": 0.0005026,
135
+ "loss": 4.270945739746094,
136
+ "step": 360
137
+ },
138
+ {
139
+ "epoch": 0.02561509942703067,
140
+ "grad_norm": 0.75,
141
+ "learning_rate": 0.0005306,
142
+ "loss": 4.125864028930664,
143
+ "step": 380
144
+ },
145
+ {
146
+ "epoch": 0.026963262554769128,
147
+ "grad_norm": 0.5625,
148
+ "learning_rate": 0.0005586,
149
+ "loss": 3.992333984375,
150
+ "step": 400
151
+ },
152
+ {
153
+ "epoch": 0.028311425682507583,
154
+ "grad_norm": 0.53515625,
155
+ "learning_rate": 0.0005866,
156
+ "loss": 3.8864391326904295,
157
+ "step": 420
158
+ },
159
+ {
160
+ "epoch": 0.02965958881024604,
161
+ "grad_norm": 0.86328125,
162
+ "learning_rate": 0.0006146,
163
+ "loss": 3.819641876220703,
164
+ "step": 440
165
+ },
166
+ {
167
+ "epoch": 0.031007751937984496,
168
+ "grad_norm": 0.7109375,
169
+ "learning_rate": 0.0006426,
170
+ "loss": 3.7153347015380858,
171
+ "step": 460
172
+ },
173
+ {
174
+ "epoch": 0.032355915065722954,
175
+ "grad_norm": 0.53515625,
176
+ "learning_rate": 0.0006705999999999999,
177
+ "loss": 3.6391063690185548,
178
+ "step": 480
179
+ },
180
+ {
181
+ "epoch": 0.03370407819346141,
182
+ "grad_norm": 0.59765625,
183
+ "learning_rate": 0.0006986,
184
+ "loss": 3.5594459533691407,
185
+ "step": 500
186
+ },
187
+ {
188
+ "epoch": 0.03505224132119986,
189
+ "grad_norm": 0.4296875,
190
+ "learning_rate": 0.0007,
191
+ "loss": 3.493515396118164,
192
+ "step": 520
193
+ },
194
+ {
195
+ "epoch": 0.03640040444893832,
196
+ "grad_norm": 0.640625,
197
+ "learning_rate": 0.0007,
198
+ "loss": 3.41654052734375,
199
+ "step": 540
200
+ },
201
+ {
202
+ "epoch": 0.03774856757667678,
203
+ "grad_norm": 0.484375,
204
+ "learning_rate": 0.0007,
205
+ "loss": 3.366413879394531,
206
+ "step": 560
207
+ },
208
+ {
209
+ "epoch": 0.03909673070441524,
210
+ "grad_norm": 0.423828125,
211
+ "learning_rate": 0.0007,
212
+ "loss": 3.3078380584716798,
213
+ "step": 580
214
+ },
215
+ {
216
+ "epoch": 0.04044489383215369,
217
+ "grad_norm": 0.423828125,
218
+ "learning_rate": 0.0007,
219
+ "loss": 3.2723438262939455,
220
+ "step": 600
221
+ },
222
+ {
223
+ "epoch": 0.04179305695989215,
224
+ "grad_norm": 0.431640625,
225
+ "learning_rate": 0.0007,
226
+ "loss": 3.2081642150878906,
227
+ "step": 620
228
+ },
229
+ {
230
+ "epoch": 0.043141220087630605,
231
+ "grad_norm": 0.44921875,
232
+ "learning_rate": 0.0007,
233
+ "loss": 3.1861698150634767,
234
+ "step": 640
235
+ },
236
+ {
237
+ "epoch": 0.044489383215369056,
238
+ "grad_norm": 0.40625,
239
+ "learning_rate": 0.0007,
240
+ "loss": 3.126705360412598,
241
+ "step": 660
242
+ },
243
+ {
244
+ "epoch": 0.045837546343107514,
245
+ "grad_norm": 0.423828125,
246
+ "learning_rate": 0.0007,
247
+ "loss": 3.1009654998779297,
248
+ "step": 680
249
+ },
250
+ {
251
+ "epoch": 0.04718570947084597,
252
+ "grad_norm": 0.41015625,
253
+ "learning_rate": 0.0007,
254
+ "loss": 3.0569204330444335,
255
+ "step": 700
256
+ },
257
+ {
258
+ "epoch": 0.04853387259858443,
259
+ "grad_norm": 0.390625,
260
+ "learning_rate": 0.0007,
261
+ "loss": 3.0219058990478516,
262
+ "step": 720
263
+ },
264
+ {
265
+ "epoch": 0.04988203572632288,
266
+ "grad_norm": 0.447265625,
267
+ "learning_rate": 0.0007,
268
+ "loss": 2.9834646224975585,
269
+ "step": 740
270
+ },
271
+ {
272
+ "epoch": 0.05123019885406134,
273
+ "grad_norm": 0.39453125,
274
+ "learning_rate": 0.0007,
275
+ "loss": 2.9665781021118165,
276
+ "step": 760
277
+ },
278
+ {
279
+ "epoch": 0.0525783619817998,
280
+ "grad_norm": 0.43359375,
281
+ "learning_rate": 0.0007,
282
+ "loss": 2.9737390518188476,
283
+ "step": 780
284
+ },
285
+ {
286
+ "epoch": 0.053926525109538256,
287
+ "grad_norm": 0.41015625,
288
+ "learning_rate": 0.0007,
289
+ "loss": 2.9195796966552736,
290
+ "step": 800
291
+ },
292
+ {
293
+ "epoch": 0.05527468823727671,
294
+ "grad_norm": 0.41015625,
295
+ "learning_rate": 0.0007,
296
+ "loss": 2.9032770156860352,
297
+ "step": 820
298
+ },
299
+ {
300
+ "epoch": 0.056622851365015166,
301
+ "grad_norm": 0.4296875,
302
+ "learning_rate": 0.0007,
303
+ "loss": 2.8703426361083983,
304
+ "step": 840
305
+ },
306
+ {
307
+ "epoch": 0.057971014492753624,
308
+ "grad_norm": 0.384765625,
309
+ "learning_rate": 0.0007,
310
+ "loss": 2.8253679275512695,
311
+ "step": 860
312
+ },
313
+ {
314
+ "epoch": 0.05931917762049208,
315
+ "grad_norm": 0.390625,
316
+ "learning_rate": 0.0007,
317
+ "loss": 2.802602195739746,
318
+ "step": 880
319
+ },
320
+ {
321
+ "epoch": 0.06066734074823053,
322
+ "grad_norm": 0.40625,
323
+ "learning_rate": 0.0007,
324
+ "loss": 2.7884506225585937,
325
+ "step": 900
326
+ },
327
+ {
328
+ "epoch": 0.06201550387596899,
329
+ "grad_norm": 0.3828125,
330
+ "learning_rate": 0.0007,
331
+ "loss": 2.7588369369506838,
332
+ "step": 920
333
+ },
334
+ {
335
+ "epoch": 0.06336366700370745,
336
+ "grad_norm": 0.41796875,
337
+ "learning_rate": 0.0007,
338
+ "loss": 2.7435585021972657,
339
+ "step": 940
340
+ },
341
+ {
342
+ "epoch": 0.06471183013144591,
343
+ "grad_norm": 0.44921875,
344
+ "learning_rate": 0.0007,
345
+ "loss": 2.717236328125,
346
+ "step": 960
347
+ },
348
+ {
349
+ "epoch": 0.06605999325918437,
350
+ "grad_norm": 0.373046875,
351
+ "learning_rate": 0.0007,
352
+ "loss": 2.699039268493652,
353
+ "step": 980
354
+ },
355
+ {
356
+ "epoch": 0.06740815638692282,
357
+ "grad_norm": 0.41015625,
358
+ "learning_rate": 0.0007,
359
+ "loss": 2.697044563293457,
360
+ "step": 1000
361
+ },
362
+ {
363
+ "epoch": 0.06875631951466127,
364
+ "grad_norm": 0.361328125,
365
+ "learning_rate": 0.0007,
366
+ "loss": 2.6641401290893554,
367
+ "step": 1020
368
+ },
369
+ {
370
+ "epoch": 0.07010448264239973,
371
+ "grad_norm": 0.376953125,
372
+ "learning_rate": 0.0007,
373
+ "loss": 2.6494163513183593,
374
+ "step": 1040
375
+ },
376
+ {
377
+ "epoch": 0.07145264577013818,
378
+ "grad_norm": 0.416015625,
379
+ "learning_rate": 0.0007,
380
+ "loss": 2.631260299682617,
381
+ "step": 1060
382
+ },
383
+ {
384
+ "epoch": 0.07280080889787664,
385
+ "grad_norm": 0.37890625,
386
+ "learning_rate": 0.0007,
387
+ "loss": 2.612077331542969,
388
+ "step": 1080
389
+ },
390
+ {
391
+ "epoch": 0.0741489720256151,
392
+ "grad_norm": 0.40234375,
393
+ "learning_rate": 0.0007,
394
+ "loss": 2.606185722351074,
395
+ "step": 1100
396
+ },
397
+ {
398
+ "epoch": 0.07549713515335356,
399
+ "grad_norm": 0.388671875,
400
+ "learning_rate": 0.0007,
401
+ "loss": 2.570285606384277,
402
+ "step": 1120
403
+ },
404
+ {
405
+ "epoch": 0.07684529828109202,
406
+ "grad_norm": 0.41796875,
407
+ "learning_rate": 0.0007,
408
+ "loss": 2.557082176208496,
409
+ "step": 1140
410
+ },
411
+ {
412
+ "epoch": 0.07819346140883048,
413
+ "grad_norm": 0.37890625,
414
+ "learning_rate": 0.0007,
415
+ "loss": 2.5526586532592774,
416
+ "step": 1160
417
+ },
418
+ {
419
+ "epoch": 0.07954162453656892,
420
+ "grad_norm": 0.388671875,
421
+ "learning_rate": 0.0007,
422
+ "loss": 2.547346305847168,
423
+ "step": 1180
424
+ },
425
+ {
426
+ "epoch": 0.08088978766430738,
427
+ "grad_norm": 0.3828125,
428
+ "learning_rate": 0.0007,
429
+ "loss": 2.5366180419921873,
430
+ "step": 1200
431
+ },
432
+ {
433
+ "epoch": 0.08223795079204584,
434
+ "grad_norm": 0.359375,
435
+ "learning_rate": 0.0007,
436
+ "loss": 2.5206745147705076,
437
+ "step": 1220
438
+ },
439
+ {
440
+ "epoch": 0.0835861139197843,
441
+ "grad_norm": 0.36328125,
442
+ "learning_rate": 0.0007,
443
+ "loss": 2.506051254272461,
444
+ "step": 1240
445
+ },
446
+ {
447
+ "epoch": 0.08493427704752275,
448
+ "grad_norm": 0.376953125,
449
+ "learning_rate": 0.0007,
450
+ "loss": 2.494335174560547,
451
+ "step": 1260
452
+ },
453
+ {
454
+ "epoch": 0.08628244017526121,
455
+ "grad_norm": 0.384765625,
456
+ "learning_rate": 0.0007,
457
+ "loss": 2.4746730804443358,
458
+ "step": 1280
459
+ },
460
+ {
461
+ "epoch": 0.08763060330299967,
462
+ "grad_norm": 0.384765625,
463
+ "learning_rate": 0.0007,
464
+ "loss": 2.4673040390014647,
465
+ "step": 1300
466
+ },
467
+ {
468
+ "epoch": 0.08897876643073811,
469
+ "grad_norm": 0.359375,
470
+ "learning_rate": 0.0007,
471
+ "loss": 2.458811950683594,
472
+ "step": 1320
473
+ },
474
+ {
475
+ "epoch": 0.09032692955847657,
476
+ "grad_norm": 0.40234375,
477
+ "learning_rate": 0.0007,
478
+ "loss": 2.444304275512695,
479
+ "step": 1340
480
+ },
481
+ {
482
+ "epoch": 0.09167509268621503,
483
+ "grad_norm": 0.3984375,
484
+ "learning_rate": 0.0007,
485
+ "loss": 2.4439945220947266,
486
+ "step": 1360
487
+ },
488
+ {
489
+ "epoch": 0.09302325581395349,
490
+ "grad_norm": 0.341796875,
491
+ "learning_rate": 0.0007,
492
+ "loss": 2.4202173233032225,
493
+ "step": 1380
494
+ },
495
+ {
496
+ "epoch": 0.09437141894169195,
497
+ "grad_norm": 0.36328125,
498
+ "learning_rate": 0.0007,
499
+ "loss": 2.4103469848632812,
500
+ "step": 1400
501
+ },
502
+ {
503
+ "epoch": 0.0957195820694304,
504
+ "grad_norm": 0.39453125,
505
+ "learning_rate": 0.0007,
506
+ "loss": 2.4228416442871095,
507
+ "step": 1420
508
+ },
509
+ {
510
+ "epoch": 0.09706774519716886,
511
+ "grad_norm": 0.3515625,
512
+ "learning_rate": 0.0007,
513
+ "loss": 2.3987619400024416,
514
+ "step": 1440
515
+ },
516
+ {
517
+ "epoch": 0.09841590832490732,
518
+ "grad_norm": 0.38671875,
519
+ "learning_rate": 0.0007,
520
+ "loss": 2.3866939544677734,
521
+ "step": 1460
522
+ },
523
+ {
524
+ "epoch": 0.09976407145264576,
525
+ "grad_norm": 0.359375,
526
+ "learning_rate": 0.0007,
527
+ "loss": 2.3730697631835938,
528
+ "step": 1480
529
+ },
530
+ {
531
+ "epoch": 0.10111223458038422,
532
+ "grad_norm": 0.359375,
533
+ "learning_rate": 0.0007,
534
+ "loss": 2.3593896865844726,
535
+ "step": 1500
536
+ },
537
+ {
538
+ "epoch": 0.10246039770812268,
539
+ "grad_norm": 0.37890625,
540
+ "learning_rate": 0.0007,
541
+ "loss": 2.362796401977539,
542
+ "step": 1520
543
+ },
544
+ {
545
+ "epoch": 0.10380856083586114,
546
+ "grad_norm": 0.3828125,
547
+ "learning_rate": 0.0007,
548
+ "loss": 2.3507095336914063,
549
+ "step": 1540
550
+ },
551
+ {
552
+ "epoch": 0.1051567239635996,
553
+ "grad_norm": 0.37890625,
554
+ "learning_rate": 0.0007,
555
+ "loss": 2.339781951904297,
556
+ "step": 1560
557
+ },
558
+ {
559
+ "epoch": 0.10650488709133805,
560
+ "grad_norm": 0.353515625,
561
+ "learning_rate": 0.0007,
562
+ "loss": 2.356917953491211,
563
+ "step": 1580
564
+ },
565
+ {
566
+ "epoch": 0.10785305021907651,
567
+ "grad_norm": 0.392578125,
568
+ "learning_rate": 0.0007,
569
+ "loss": 2.325514030456543,
570
+ "step": 1600
571
+ },
572
+ {
573
+ "epoch": 0.10920121334681497,
574
+ "grad_norm": 0.33984375,
575
+ "learning_rate": 0.0007,
576
+ "loss": 2.3245080947875976,
577
+ "step": 1620
578
+ },
579
+ {
580
+ "epoch": 0.11054937647455342,
581
+ "grad_norm": 0.349609375,
582
+ "learning_rate": 0.0007,
583
+ "loss": 2.3359960556030273,
584
+ "step": 1640
585
+ },
586
+ {
587
+ "epoch": 0.11189753960229187,
588
+ "grad_norm": 0.3359375,
589
+ "learning_rate": 0.0007,
590
+ "loss": 2.3171085357666015,
591
+ "step": 1660
592
+ },
593
+ {
594
+ "epoch": 0.11324570273003033,
595
+ "grad_norm": 0.37890625,
596
+ "learning_rate": 0.0007,
597
+ "loss": 2.299543571472168,
598
+ "step": 1680
599
+ },
600
+ {
601
+ "epoch": 0.11459386585776879,
602
+ "grad_norm": 0.3515625,
603
+ "learning_rate": 0.0007,
604
+ "loss": 2.309956932067871,
605
+ "step": 1700
606
+ },
607
+ {
608
+ "epoch": 0.11594202898550725,
609
+ "grad_norm": 0.3828125,
610
+ "learning_rate": 0.0007,
611
+ "loss": 2.2854846954345702,
612
+ "step": 1720
613
+ },
614
+ {
615
+ "epoch": 0.1172901921132457,
616
+ "grad_norm": 0.349609375,
617
+ "learning_rate": 0.0007,
618
+ "loss": 2.2760204315185546,
619
+ "step": 1740
620
+ },
621
+ {
622
+ "epoch": 0.11863835524098416,
623
+ "grad_norm": 0.3515625,
624
+ "learning_rate": 0.0007,
625
+ "loss": 2.279789924621582,
626
+ "step": 1760
627
+ },
628
+ {
629
+ "epoch": 0.11998651836872262,
630
+ "grad_norm": 0.3671875,
631
+ "learning_rate": 0.0007,
632
+ "loss": 2.2942825317382813,
633
+ "step": 1780
634
+ },
635
+ {
636
+ "epoch": 0.12133468149646107,
637
+ "grad_norm": 0.341796875,
638
+ "learning_rate": 0.0007,
639
+ "loss": 2.2891586303710936,
640
+ "step": 1800
641
+ },
642
+ {
643
+ "epoch": 0.12268284462419952,
644
+ "grad_norm": 0.3828125,
645
+ "learning_rate": 0.0007,
646
+ "loss": 2.279863166809082,
647
+ "step": 1820
648
+ },
649
+ {
650
+ "epoch": 0.12403100775193798,
651
+ "grad_norm": 0.359375,
652
+ "learning_rate": 0.0007,
653
+ "loss": 2.2473472595214843,
654
+ "step": 1840
655
+ },
656
+ {
657
+ "epoch": 0.12537917087967643,
658
+ "grad_norm": 0.33984375,
659
+ "learning_rate": 0.0007,
660
+ "loss": 2.248352813720703,
661
+ "step": 1860
662
+ },
663
+ {
664
+ "epoch": 0.1267273340074149,
665
+ "grad_norm": 0.322265625,
666
+ "learning_rate": 0.0007,
667
+ "loss": 2.239566612243652,
668
+ "step": 1880
669
+ },
670
+ {
671
+ "epoch": 0.12807549713515334,
672
+ "grad_norm": 0.3359375,
673
+ "learning_rate": 0.0007,
674
+ "loss": 2.2460857391357423,
675
+ "step": 1900
676
+ },
677
+ {
678
+ "epoch": 0.12942366026289182,
679
+ "grad_norm": 0.345703125,
680
+ "learning_rate": 0.0007,
681
+ "loss": 2.222591209411621,
682
+ "step": 1920
683
+ },
684
+ {
685
+ "epoch": 0.13077182339063026,
686
+ "grad_norm": 0.35546875,
687
+ "learning_rate": 0.0007,
688
+ "loss": 2.2339385986328124,
689
+ "step": 1940
690
+ },
691
+ {
692
+ "epoch": 0.13211998651836873,
693
+ "grad_norm": 0.337890625,
694
+ "learning_rate": 0.0007,
695
+ "loss": 2.2090906143188476,
696
+ "step": 1960
697
+ },
698
+ {
699
+ "epoch": 0.13346814964610718,
700
+ "grad_norm": 0.36328125,
701
+ "learning_rate": 0.0007,
702
+ "loss": 2.2296125411987306,
703
+ "step": 1980
704
+ },
705
+ {
706
+ "epoch": 0.13481631277384565,
707
+ "grad_norm": 0.345703125,
708
+ "learning_rate": 0.0007,
709
+ "loss": 2.20355110168457,
710
+ "step": 2000
711
+ },
712
+ {
713
+ "epoch": 0.1361644759015841,
714
+ "grad_norm": 0.34765625,
715
+ "learning_rate": 0.0007,
716
+ "loss": 2.211279296875,
717
+ "step": 2020
718
+ },
719
+ {
720
+ "epoch": 0.13751263902932254,
721
+ "grad_norm": 0.37890625,
722
+ "learning_rate": 0.0007,
723
+ "loss": 2.1930706024169924,
724
+ "step": 2040
725
+ },
726
+ {
727
+ "epoch": 0.138860802157061,
728
+ "grad_norm": 0.3359375,
729
+ "learning_rate": 0.0007,
730
+ "loss": 2.211184501647949,
731
+ "step": 2060
732
+ },
733
+ {
734
+ "epoch": 0.14020896528479945,
735
+ "grad_norm": 0.333984375,
736
+ "learning_rate": 0.0007,
737
+ "loss": 2.1871965408325194,
738
+ "step": 2080
739
+ },
740
+ {
741
+ "epoch": 0.14155712841253792,
742
+ "grad_norm": 0.33203125,
743
+ "learning_rate": 0.0007,
744
+ "loss": 2.178455352783203,
745
+ "step": 2100
746
+ },
747
+ {
748
+ "epoch": 0.14290529154027637,
749
+ "grad_norm": 0.35546875,
750
+ "learning_rate": 0.0007,
751
+ "loss": 2.186496925354004,
752
+ "step": 2120
753
+ },
754
+ {
755
+ "epoch": 0.14425345466801484,
756
+ "grad_norm": 0.341796875,
757
+ "learning_rate": 0.0007,
758
+ "loss": 2.1872365951538084,
759
+ "step": 2140
760
+ },
761
+ {
762
+ "epoch": 0.14560161779575329,
763
+ "grad_norm": 0.337890625,
764
+ "learning_rate": 0.0007,
765
+ "loss": 2.1733598709106445,
766
+ "step": 2160
767
+ },
768
+ {
769
+ "epoch": 0.14694978092349173,
770
+ "grad_norm": 0.3359375,
771
+ "learning_rate": 0.0007,
772
+ "loss": 2.1814584732055664,
773
+ "step": 2180
774
+ },
775
+ {
776
+ "epoch": 0.1482979440512302,
777
+ "grad_norm": 0.328125,
778
+ "learning_rate": 0.0007,
779
+ "loss": 2.157553291320801,
780
+ "step": 2200
781
+ },
782
+ {
783
+ "epoch": 0.14964610717896865,
784
+ "grad_norm": 0.341796875,
785
+ "learning_rate": 0.0007,
786
+ "loss": 2.1652835845947265,
787
+ "step": 2220
788
+ },
789
+ {
790
+ "epoch": 0.15099427030670712,
791
+ "grad_norm": 0.34375,
792
+ "learning_rate": 0.0007,
793
+ "loss": 2.1730274200439452,
794
+ "step": 2240
795
+ },
796
+ {
797
+ "epoch": 0.15234243343444556,
798
+ "grad_norm": 0.353515625,
799
+ "learning_rate": 0.0007,
800
+ "loss": 2.1498680114746094,
801
+ "step": 2260
802
+ },
803
+ {
804
+ "epoch": 0.15369059656218403,
805
+ "grad_norm": 0.326171875,
806
+ "learning_rate": 0.0007,
807
+ "loss": 2.1618186950683596,
808
+ "step": 2280
809
+ },
810
+ {
811
+ "epoch": 0.15503875968992248,
812
+ "grad_norm": 0.337890625,
813
+ "learning_rate": 0.0007,
814
+ "loss": 2.142350959777832,
815
+ "step": 2300
816
+ }
817
+ ],
818
+ "logging_steps": 20,
819
+ "max_steps": 2500,
820
+ "num_input_tokens_seen": 0,
821
+ "num_train_epochs": 1,
822
+ "save_steps": 100,
823
+ "stateful_callbacks": {
824
+ "TrainerControl": {
825
+ "args": {
826
+ "should_epoch_stop": false,
827
+ "should_evaluate": false,
828
+ "should_log": false,
829
+ "should_save": true,
830
+ "should_training_stop": false
831
+ },
832
+ "attributes": {}
833
+ }
834
+ },
835
+ "total_flos": 7420452746035200.0,
836
+ "train_batch_size": 64,
837
+ "trial_name": null,
838
+ "trial_params": null
839
+ }
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2300/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:efb8cca2a9ea6ff541b7562ec4e2e31a24b5072ef5136a6523ae9f9c0012caa0
3
+ size 4920
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "situglu_low",
3
+ "architectures": [
4
+ "TinyLlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 2,
11
+ "head_dim": 32,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 128,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 256,
16
+ "max_position_embeddings": 512,
17
+ "mlp_bias": false,
18
+ "mlp_type": "glu",
19
+ "model_type": "tiny_llama",
20
+ "num_attention_heads": 4,
21
+ "num_hidden_layers": 100,
22
+ "num_key_value_heads": 4,
23
+ "pad_token_id": 0,
24
+ "pretraining_tp": 1,
25
+ "rms_norm_eps": 1e-06,
26
+ "rope_parameters": {
27
+ "rope_theta": 10000.0,
28
+ "rope_type": "default"
29
+ },
30
+ "tie_word_embeddings": true,
31
+ "tokenizer_name": "w-ahmad/tiny-stories-tokenizer",
32
+ "transformers_version": "5.16.0.dev0",
33
+ "use_cache": false,
34
+ "vocab_size": 4096,
35
+ "waleed_beta": 10.0
36
+ }
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8600b8a0c74c737ad96f20f7a55fbf5c7f76d052cf9bbff102242424174336a3
3
+ size 33967272
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3cf804408a49a11bc988b20027630b93fe96e9a249fa4d8d91dad83026763de4
3
+ size 68504996
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9d9cd6a0487226e5bd30d1846894c82af483733ab4381b75bae9c0745e05d405
3
+ size 14244
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cffe3454608d7821f042822654b62bd9020d78872ebdecde4f22f94320596378
3
+ size 1064
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/tokenizer_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|endoftext|>",
5
+ "eos_token": "<|endoftext|>",
6
+ "errors": "replace",
7
+ "is_local": false,
8
+ "local_files_only": false,
9
+ "model_max_length": 1024,
10
+ "pad_token": "<|endoftext|>",
11
+ "tokenizer_class": "GPT2Tokenizer",
12
+ "unk_token": "<|endoftext|>"
13
+ }
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/trainer_state.json ADDED
@@ -0,0 +1,874 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.16177957532861476,
6
+ "eval_steps": 2498,
7
+ "global_step": 2400,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0013481631277384564,
14
+ "grad_norm": 2.078125,
15
+ "learning_rate": 2.66e-05,
16
+ "loss": 8.299667358398438,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.3515625,
22
+ "learning_rate": 5.46e-05,
23
+ "loss": 8.047861480712891,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 1.2734375,
29
+ "learning_rate": 8.259999999999999e-05,
30
+ "loss": 7.838032531738281,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 1.265625,
36
+ "learning_rate": 0.0001106,
37
+ "loss": 7.572964477539062,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 1.1953125,
43
+ "learning_rate": 0.0001386,
44
+ "loss": 7.233274841308594,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 1.1328125,
50
+ "learning_rate": 0.00016659999999999998,
51
+ "loss": 6.860704803466797,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 0.94921875,
57
+ "learning_rate": 0.00019460000000000001,
58
+ "loss": 6.517856597900391,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 0.8046875,
64
+ "learning_rate": 0.0002226,
65
+ "loss": 6.239397811889648,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 0.65234375,
71
+ "learning_rate": 0.00025059999999999997,
72
+ "loss": 6.039968109130859,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 0.5625,
78
+ "learning_rate": 0.0002786,
79
+ "loss": 5.863864517211914,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.01482979440512302,
84
+ "grad_norm": 0.67578125,
85
+ "learning_rate": 0.00030659999999999997,
86
+ "loss": 5.679952239990234,
87
+ "step": 220
88
+ },
89
+ {
90
+ "epoch": 0.016177957532861477,
91
+ "grad_norm": 1.484375,
92
+ "learning_rate": 0.0003346,
93
+ "loss": 5.470568084716797,
94
+ "step": 240
95
+ },
96
+ {
97
+ "epoch": 0.01752612066059993,
98
+ "grad_norm": 1.9765625,
99
+ "learning_rate": 0.00036260000000000003,
100
+ "loss": 5.266762542724609,
101
+ "step": 260
102
+ },
103
+ {
104
+ "epoch": 0.01887428378833839,
105
+ "grad_norm": 0.6015625,
106
+ "learning_rate": 0.0003906,
107
+ "loss": 5.04127197265625,
108
+ "step": 280
109
+ },
110
+ {
111
+ "epoch": 0.020222446916076844,
112
+ "grad_norm": 1.0859375,
113
+ "learning_rate": 0.0004186,
114
+ "loss": 4.803184509277344,
115
+ "step": 300
116
+ },
117
+ {
118
+ "epoch": 0.021570610043815303,
119
+ "grad_norm": 1.15625,
120
+ "learning_rate": 0.0004466,
121
+ "loss": 4.589862823486328,
122
+ "step": 320
123
+ },
124
+ {
125
+ "epoch": 0.022918773171553757,
126
+ "grad_norm": 0.6953125,
127
+ "learning_rate": 0.00047460000000000004,
128
+ "loss": 4.416217803955078,
129
+ "step": 340
130
+ },
131
+ {
132
+ "epoch": 0.024266936299292215,
133
+ "grad_norm": 0.8046875,
134
+ "learning_rate": 0.0005026,
135
+ "loss": 4.270945739746094,
136
+ "step": 360
137
+ },
138
+ {
139
+ "epoch": 0.02561509942703067,
140
+ "grad_norm": 0.75,
141
+ "learning_rate": 0.0005306,
142
+ "loss": 4.125864028930664,
143
+ "step": 380
144
+ },
145
+ {
146
+ "epoch": 0.026963262554769128,
147
+ "grad_norm": 0.5625,
148
+ "learning_rate": 0.0005586,
149
+ "loss": 3.992333984375,
150
+ "step": 400
151
+ },
152
+ {
153
+ "epoch": 0.028311425682507583,
154
+ "grad_norm": 0.53515625,
155
+ "learning_rate": 0.0005866,
156
+ "loss": 3.8864391326904295,
157
+ "step": 420
158
+ },
159
+ {
160
+ "epoch": 0.02965958881024604,
161
+ "grad_norm": 0.86328125,
162
+ "learning_rate": 0.0006146,
163
+ "loss": 3.819641876220703,
164
+ "step": 440
165
+ },
166
+ {
167
+ "epoch": 0.031007751937984496,
168
+ "grad_norm": 0.7109375,
169
+ "learning_rate": 0.0006426,
170
+ "loss": 3.7153347015380858,
171
+ "step": 460
172
+ },
173
+ {
174
+ "epoch": 0.032355915065722954,
175
+ "grad_norm": 0.53515625,
176
+ "learning_rate": 0.0006705999999999999,
177
+ "loss": 3.6391063690185548,
178
+ "step": 480
179
+ },
180
+ {
181
+ "epoch": 0.03370407819346141,
182
+ "grad_norm": 0.59765625,
183
+ "learning_rate": 0.0006986,
184
+ "loss": 3.5594459533691407,
185
+ "step": 500
186
+ },
187
+ {
188
+ "epoch": 0.03505224132119986,
189
+ "grad_norm": 0.4296875,
190
+ "learning_rate": 0.0007,
191
+ "loss": 3.493515396118164,
192
+ "step": 520
193
+ },
194
+ {
195
+ "epoch": 0.03640040444893832,
196
+ "grad_norm": 0.640625,
197
+ "learning_rate": 0.0007,
198
+ "loss": 3.41654052734375,
199
+ "step": 540
200
+ },
201
+ {
202
+ "epoch": 0.03774856757667678,
203
+ "grad_norm": 0.484375,
204
+ "learning_rate": 0.0007,
205
+ "loss": 3.366413879394531,
206
+ "step": 560
207
+ },
208
+ {
209
+ "epoch": 0.03909673070441524,
210
+ "grad_norm": 0.423828125,
211
+ "learning_rate": 0.0007,
212
+ "loss": 3.3078380584716798,
213
+ "step": 580
214
+ },
215
+ {
216
+ "epoch": 0.04044489383215369,
217
+ "grad_norm": 0.423828125,
218
+ "learning_rate": 0.0007,
219
+ "loss": 3.2723438262939455,
220
+ "step": 600
221
+ },
222
+ {
223
+ "epoch": 0.04179305695989215,
224
+ "grad_norm": 0.431640625,
225
+ "learning_rate": 0.0007,
226
+ "loss": 3.2081642150878906,
227
+ "step": 620
228
+ },
229
+ {
230
+ "epoch": 0.043141220087630605,
231
+ "grad_norm": 0.44921875,
232
+ "learning_rate": 0.0007,
233
+ "loss": 3.1861698150634767,
234
+ "step": 640
235
+ },
236
+ {
237
+ "epoch": 0.044489383215369056,
238
+ "grad_norm": 0.40625,
239
+ "learning_rate": 0.0007,
240
+ "loss": 3.126705360412598,
241
+ "step": 660
242
+ },
243
+ {
244
+ "epoch": 0.045837546343107514,
245
+ "grad_norm": 0.423828125,
246
+ "learning_rate": 0.0007,
247
+ "loss": 3.1009654998779297,
248
+ "step": 680
249
+ },
250
+ {
251
+ "epoch": 0.04718570947084597,
252
+ "grad_norm": 0.41015625,
253
+ "learning_rate": 0.0007,
254
+ "loss": 3.0569204330444335,
255
+ "step": 700
256
+ },
257
+ {
258
+ "epoch": 0.04853387259858443,
259
+ "grad_norm": 0.390625,
260
+ "learning_rate": 0.0007,
261
+ "loss": 3.0219058990478516,
262
+ "step": 720
263
+ },
264
+ {
265
+ "epoch": 0.04988203572632288,
266
+ "grad_norm": 0.447265625,
267
+ "learning_rate": 0.0007,
268
+ "loss": 2.9834646224975585,
269
+ "step": 740
270
+ },
271
+ {
272
+ "epoch": 0.05123019885406134,
273
+ "grad_norm": 0.39453125,
274
+ "learning_rate": 0.0007,
275
+ "loss": 2.9665781021118165,
276
+ "step": 760
277
+ },
278
+ {
279
+ "epoch": 0.0525783619817998,
280
+ "grad_norm": 0.43359375,
281
+ "learning_rate": 0.0007,
282
+ "loss": 2.9737390518188476,
283
+ "step": 780
284
+ },
285
+ {
286
+ "epoch": 0.053926525109538256,
287
+ "grad_norm": 0.41015625,
288
+ "learning_rate": 0.0007,
289
+ "loss": 2.9195796966552736,
290
+ "step": 800
291
+ },
292
+ {
293
+ "epoch": 0.05527468823727671,
294
+ "grad_norm": 0.41015625,
295
+ "learning_rate": 0.0007,
296
+ "loss": 2.9032770156860352,
297
+ "step": 820
298
+ },
299
+ {
300
+ "epoch": 0.056622851365015166,
301
+ "grad_norm": 0.4296875,
302
+ "learning_rate": 0.0007,
303
+ "loss": 2.8703426361083983,
304
+ "step": 840
305
+ },
306
+ {
307
+ "epoch": 0.057971014492753624,
308
+ "grad_norm": 0.384765625,
309
+ "learning_rate": 0.0007,
310
+ "loss": 2.8253679275512695,
311
+ "step": 860
312
+ },
313
+ {
314
+ "epoch": 0.05931917762049208,
315
+ "grad_norm": 0.390625,
316
+ "learning_rate": 0.0007,
317
+ "loss": 2.802602195739746,
318
+ "step": 880
319
+ },
320
+ {
321
+ "epoch": 0.06066734074823053,
322
+ "grad_norm": 0.40625,
323
+ "learning_rate": 0.0007,
324
+ "loss": 2.7884506225585937,
325
+ "step": 900
326
+ },
327
+ {
328
+ "epoch": 0.06201550387596899,
329
+ "grad_norm": 0.3828125,
330
+ "learning_rate": 0.0007,
331
+ "loss": 2.7588369369506838,
332
+ "step": 920
333
+ },
334
+ {
335
+ "epoch": 0.06336366700370745,
336
+ "grad_norm": 0.41796875,
337
+ "learning_rate": 0.0007,
338
+ "loss": 2.7435585021972657,
339
+ "step": 940
340
+ },
341
+ {
342
+ "epoch": 0.06471183013144591,
343
+ "grad_norm": 0.44921875,
344
+ "learning_rate": 0.0007,
345
+ "loss": 2.717236328125,
346
+ "step": 960
347
+ },
348
+ {
349
+ "epoch": 0.06605999325918437,
350
+ "grad_norm": 0.373046875,
351
+ "learning_rate": 0.0007,
352
+ "loss": 2.699039268493652,
353
+ "step": 980
354
+ },
355
+ {
356
+ "epoch": 0.06740815638692282,
357
+ "grad_norm": 0.41015625,
358
+ "learning_rate": 0.0007,
359
+ "loss": 2.697044563293457,
360
+ "step": 1000
361
+ },
362
+ {
363
+ "epoch": 0.06875631951466127,
364
+ "grad_norm": 0.361328125,
365
+ "learning_rate": 0.0007,
366
+ "loss": 2.6641401290893554,
367
+ "step": 1020
368
+ },
369
+ {
370
+ "epoch": 0.07010448264239973,
371
+ "grad_norm": 0.376953125,
372
+ "learning_rate": 0.0007,
373
+ "loss": 2.6494163513183593,
374
+ "step": 1040
375
+ },
376
+ {
377
+ "epoch": 0.07145264577013818,
378
+ "grad_norm": 0.416015625,
379
+ "learning_rate": 0.0007,
380
+ "loss": 2.631260299682617,
381
+ "step": 1060
382
+ },
383
+ {
384
+ "epoch": 0.07280080889787664,
385
+ "grad_norm": 0.37890625,
386
+ "learning_rate": 0.0007,
387
+ "loss": 2.612077331542969,
388
+ "step": 1080
389
+ },
390
+ {
391
+ "epoch": 0.0741489720256151,
392
+ "grad_norm": 0.40234375,
393
+ "learning_rate": 0.0007,
394
+ "loss": 2.606185722351074,
395
+ "step": 1100
396
+ },
397
+ {
398
+ "epoch": 0.07549713515335356,
399
+ "grad_norm": 0.388671875,
400
+ "learning_rate": 0.0007,
401
+ "loss": 2.570285606384277,
402
+ "step": 1120
403
+ },
404
+ {
405
+ "epoch": 0.07684529828109202,
406
+ "grad_norm": 0.41796875,
407
+ "learning_rate": 0.0007,
408
+ "loss": 2.557082176208496,
409
+ "step": 1140
410
+ },
411
+ {
412
+ "epoch": 0.07819346140883048,
413
+ "grad_norm": 0.37890625,
414
+ "learning_rate": 0.0007,
415
+ "loss": 2.5526586532592774,
416
+ "step": 1160
417
+ },
418
+ {
419
+ "epoch": 0.07954162453656892,
420
+ "grad_norm": 0.388671875,
421
+ "learning_rate": 0.0007,
422
+ "loss": 2.547346305847168,
423
+ "step": 1180
424
+ },
425
+ {
426
+ "epoch": 0.08088978766430738,
427
+ "grad_norm": 0.3828125,
428
+ "learning_rate": 0.0007,
429
+ "loss": 2.5366180419921873,
430
+ "step": 1200
431
+ },
432
+ {
433
+ "epoch": 0.08223795079204584,
434
+ "grad_norm": 0.359375,
435
+ "learning_rate": 0.0007,
436
+ "loss": 2.5206745147705076,
437
+ "step": 1220
438
+ },
439
+ {
440
+ "epoch": 0.0835861139197843,
441
+ "grad_norm": 0.36328125,
442
+ "learning_rate": 0.0007,
443
+ "loss": 2.506051254272461,
444
+ "step": 1240
445
+ },
446
+ {
447
+ "epoch": 0.08493427704752275,
448
+ "grad_norm": 0.376953125,
449
+ "learning_rate": 0.0007,
450
+ "loss": 2.494335174560547,
451
+ "step": 1260
452
+ },
453
+ {
454
+ "epoch": 0.08628244017526121,
455
+ "grad_norm": 0.384765625,
456
+ "learning_rate": 0.0007,
457
+ "loss": 2.4746730804443358,
458
+ "step": 1280
459
+ },
460
+ {
461
+ "epoch": 0.08763060330299967,
462
+ "grad_norm": 0.384765625,
463
+ "learning_rate": 0.0007,
464
+ "loss": 2.4673040390014647,
465
+ "step": 1300
466
+ },
467
+ {
468
+ "epoch": 0.08897876643073811,
469
+ "grad_norm": 0.359375,
470
+ "learning_rate": 0.0007,
471
+ "loss": 2.458811950683594,
472
+ "step": 1320
473
+ },
474
+ {
475
+ "epoch": 0.09032692955847657,
476
+ "grad_norm": 0.40234375,
477
+ "learning_rate": 0.0007,
478
+ "loss": 2.444304275512695,
479
+ "step": 1340
480
+ },
481
+ {
482
+ "epoch": 0.09167509268621503,
483
+ "grad_norm": 0.3984375,
484
+ "learning_rate": 0.0007,
485
+ "loss": 2.4439945220947266,
486
+ "step": 1360
487
+ },
488
+ {
489
+ "epoch": 0.09302325581395349,
490
+ "grad_norm": 0.341796875,
491
+ "learning_rate": 0.0007,
492
+ "loss": 2.4202173233032225,
493
+ "step": 1380
494
+ },
495
+ {
496
+ "epoch": 0.09437141894169195,
497
+ "grad_norm": 0.36328125,
498
+ "learning_rate": 0.0007,
499
+ "loss": 2.4103469848632812,
500
+ "step": 1400
501
+ },
502
+ {
503
+ "epoch": 0.0957195820694304,
504
+ "grad_norm": 0.39453125,
505
+ "learning_rate": 0.0007,
506
+ "loss": 2.4228416442871095,
507
+ "step": 1420
508
+ },
509
+ {
510
+ "epoch": 0.09706774519716886,
511
+ "grad_norm": 0.3515625,
512
+ "learning_rate": 0.0007,
513
+ "loss": 2.3987619400024416,
514
+ "step": 1440
515
+ },
516
+ {
517
+ "epoch": 0.09841590832490732,
518
+ "grad_norm": 0.38671875,
519
+ "learning_rate": 0.0007,
520
+ "loss": 2.3866939544677734,
521
+ "step": 1460
522
+ },
523
+ {
524
+ "epoch": 0.09976407145264576,
525
+ "grad_norm": 0.359375,
526
+ "learning_rate": 0.0007,
527
+ "loss": 2.3730697631835938,
528
+ "step": 1480
529
+ },
530
+ {
531
+ "epoch": 0.10111223458038422,
532
+ "grad_norm": 0.359375,
533
+ "learning_rate": 0.0007,
534
+ "loss": 2.3593896865844726,
535
+ "step": 1500
536
+ },
537
+ {
538
+ "epoch": 0.10246039770812268,
539
+ "grad_norm": 0.37890625,
540
+ "learning_rate": 0.0007,
541
+ "loss": 2.362796401977539,
542
+ "step": 1520
543
+ },
544
+ {
545
+ "epoch": 0.10380856083586114,
546
+ "grad_norm": 0.3828125,
547
+ "learning_rate": 0.0007,
548
+ "loss": 2.3507095336914063,
549
+ "step": 1540
550
+ },
551
+ {
552
+ "epoch": 0.1051567239635996,
553
+ "grad_norm": 0.37890625,
554
+ "learning_rate": 0.0007,
555
+ "loss": 2.339781951904297,
556
+ "step": 1560
557
+ },
558
+ {
559
+ "epoch": 0.10650488709133805,
560
+ "grad_norm": 0.353515625,
561
+ "learning_rate": 0.0007,
562
+ "loss": 2.356917953491211,
563
+ "step": 1580
564
+ },
565
+ {
566
+ "epoch": 0.10785305021907651,
567
+ "grad_norm": 0.392578125,
568
+ "learning_rate": 0.0007,
569
+ "loss": 2.325514030456543,
570
+ "step": 1600
571
+ },
572
+ {
573
+ "epoch": 0.10920121334681497,
574
+ "grad_norm": 0.33984375,
575
+ "learning_rate": 0.0007,
576
+ "loss": 2.3245080947875976,
577
+ "step": 1620
578
+ },
579
+ {
580
+ "epoch": 0.11054937647455342,
581
+ "grad_norm": 0.349609375,
582
+ "learning_rate": 0.0007,
583
+ "loss": 2.3359960556030273,
584
+ "step": 1640
585
+ },
586
+ {
587
+ "epoch": 0.11189753960229187,
588
+ "grad_norm": 0.3359375,
589
+ "learning_rate": 0.0007,
590
+ "loss": 2.3171085357666015,
591
+ "step": 1660
592
+ },
593
+ {
594
+ "epoch": 0.11324570273003033,
595
+ "grad_norm": 0.37890625,
596
+ "learning_rate": 0.0007,
597
+ "loss": 2.299543571472168,
598
+ "step": 1680
599
+ },
600
+ {
601
+ "epoch": 0.11459386585776879,
602
+ "grad_norm": 0.3515625,
603
+ "learning_rate": 0.0007,
604
+ "loss": 2.309956932067871,
605
+ "step": 1700
606
+ },
607
+ {
608
+ "epoch": 0.11594202898550725,
609
+ "grad_norm": 0.3828125,
610
+ "learning_rate": 0.0007,
611
+ "loss": 2.2854846954345702,
612
+ "step": 1720
613
+ },
614
+ {
615
+ "epoch": 0.1172901921132457,
616
+ "grad_norm": 0.349609375,
617
+ "learning_rate": 0.0007,
618
+ "loss": 2.2760204315185546,
619
+ "step": 1740
620
+ },
621
+ {
622
+ "epoch": 0.11863835524098416,
623
+ "grad_norm": 0.3515625,
624
+ "learning_rate": 0.0007,
625
+ "loss": 2.279789924621582,
626
+ "step": 1760
627
+ },
628
+ {
629
+ "epoch": 0.11998651836872262,
630
+ "grad_norm": 0.3671875,
631
+ "learning_rate": 0.0007,
632
+ "loss": 2.2942825317382813,
633
+ "step": 1780
634
+ },
635
+ {
636
+ "epoch": 0.12133468149646107,
637
+ "grad_norm": 0.341796875,
638
+ "learning_rate": 0.0007,
639
+ "loss": 2.2891586303710936,
640
+ "step": 1800
641
+ },
642
+ {
643
+ "epoch": 0.12268284462419952,
644
+ "grad_norm": 0.3828125,
645
+ "learning_rate": 0.0007,
646
+ "loss": 2.279863166809082,
647
+ "step": 1820
648
+ },
649
+ {
650
+ "epoch": 0.12403100775193798,
651
+ "grad_norm": 0.359375,
652
+ "learning_rate": 0.0007,
653
+ "loss": 2.2473472595214843,
654
+ "step": 1840
655
+ },
656
+ {
657
+ "epoch": 0.12537917087967643,
658
+ "grad_norm": 0.33984375,
659
+ "learning_rate": 0.0007,
660
+ "loss": 2.248352813720703,
661
+ "step": 1860
662
+ },
663
+ {
664
+ "epoch": 0.1267273340074149,
665
+ "grad_norm": 0.322265625,
666
+ "learning_rate": 0.0007,
667
+ "loss": 2.239566612243652,
668
+ "step": 1880
669
+ },
670
+ {
671
+ "epoch": 0.12807549713515334,
672
+ "grad_norm": 0.3359375,
673
+ "learning_rate": 0.0007,
674
+ "loss": 2.2460857391357423,
675
+ "step": 1900
676
+ },
677
+ {
678
+ "epoch": 0.12942366026289182,
679
+ "grad_norm": 0.345703125,
680
+ "learning_rate": 0.0007,
681
+ "loss": 2.222591209411621,
682
+ "step": 1920
683
+ },
684
+ {
685
+ "epoch": 0.13077182339063026,
686
+ "grad_norm": 0.35546875,
687
+ "learning_rate": 0.0007,
688
+ "loss": 2.2339385986328124,
689
+ "step": 1940
690
+ },
691
+ {
692
+ "epoch": 0.13211998651836873,
693
+ "grad_norm": 0.337890625,
694
+ "learning_rate": 0.0007,
695
+ "loss": 2.2090906143188476,
696
+ "step": 1960
697
+ },
698
+ {
699
+ "epoch": 0.13346814964610718,
700
+ "grad_norm": 0.36328125,
701
+ "learning_rate": 0.0007,
702
+ "loss": 2.2296125411987306,
703
+ "step": 1980
704
+ },
705
+ {
706
+ "epoch": 0.13481631277384565,
707
+ "grad_norm": 0.345703125,
708
+ "learning_rate": 0.0007,
709
+ "loss": 2.20355110168457,
710
+ "step": 2000
711
+ },
712
+ {
713
+ "epoch": 0.1361644759015841,
714
+ "grad_norm": 0.34765625,
715
+ "learning_rate": 0.0007,
716
+ "loss": 2.211279296875,
717
+ "step": 2020
718
+ },
719
+ {
720
+ "epoch": 0.13751263902932254,
721
+ "grad_norm": 0.37890625,
722
+ "learning_rate": 0.0007,
723
+ "loss": 2.1930706024169924,
724
+ "step": 2040
725
+ },
726
+ {
727
+ "epoch": 0.138860802157061,
728
+ "grad_norm": 0.3359375,
729
+ "learning_rate": 0.0007,
730
+ "loss": 2.211184501647949,
731
+ "step": 2060
732
+ },
733
+ {
734
+ "epoch": 0.14020896528479945,
735
+ "grad_norm": 0.333984375,
736
+ "learning_rate": 0.0007,
737
+ "loss": 2.1871965408325194,
738
+ "step": 2080
739
+ },
740
+ {
741
+ "epoch": 0.14155712841253792,
742
+ "grad_norm": 0.33203125,
743
+ "learning_rate": 0.0007,
744
+ "loss": 2.178455352783203,
745
+ "step": 2100
746
+ },
747
+ {
748
+ "epoch": 0.14290529154027637,
749
+ "grad_norm": 0.35546875,
750
+ "learning_rate": 0.0007,
751
+ "loss": 2.186496925354004,
752
+ "step": 2120
753
+ },
754
+ {
755
+ "epoch": 0.14425345466801484,
756
+ "grad_norm": 0.341796875,
757
+ "learning_rate": 0.0007,
758
+ "loss": 2.1872365951538084,
759
+ "step": 2140
760
+ },
761
+ {
762
+ "epoch": 0.14560161779575329,
763
+ "grad_norm": 0.337890625,
764
+ "learning_rate": 0.0007,
765
+ "loss": 2.1733598709106445,
766
+ "step": 2160
767
+ },
768
+ {
769
+ "epoch": 0.14694978092349173,
770
+ "grad_norm": 0.3359375,
771
+ "learning_rate": 0.0007,
772
+ "loss": 2.1814584732055664,
773
+ "step": 2180
774
+ },
775
+ {
776
+ "epoch": 0.1482979440512302,
777
+ "grad_norm": 0.328125,
778
+ "learning_rate": 0.0007,
779
+ "loss": 2.157553291320801,
780
+ "step": 2200
781
+ },
782
+ {
783
+ "epoch": 0.14964610717896865,
784
+ "grad_norm": 0.341796875,
785
+ "learning_rate": 0.0007,
786
+ "loss": 2.1652835845947265,
787
+ "step": 2220
788
+ },
789
+ {
790
+ "epoch": 0.15099427030670712,
791
+ "grad_norm": 0.34375,
792
+ "learning_rate": 0.0007,
793
+ "loss": 2.1730274200439452,
794
+ "step": 2240
795
+ },
796
+ {
797
+ "epoch": 0.15234243343444556,
798
+ "grad_norm": 0.353515625,
799
+ "learning_rate": 0.0007,
800
+ "loss": 2.1498680114746094,
801
+ "step": 2260
802
+ },
803
+ {
804
+ "epoch": 0.15369059656218403,
805
+ "grad_norm": 0.326171875,
806
+ "learning_rate": 0.0007,
807
+ "loss": 2.1618186950683596,
808
+ "step": 2280
809
+ },
810
+ {
811
+ "epoch": 0.15503875968992248,
812
+ "grad_norm": 0.337890625,
813
+ "learning_rate": 0.0007,
814
+ "loss": 2.142350959777832,
815
+ "step": 2300
816
+ },
817
+ {
818
+ "epoch": 0.15638692281766095,
819
+ "grad_norm": 0.333984375,
820
+ "learning_rate": 0.0007,
821
+ "loss": 2.1518985748291017,
822
+ "step": 2320
823
+ },
824
+ {
825
+ "epoch": 0.1577350859453994,
826
+ "grad_norm": 0.337890625,
827
+ "learning_rate": 0.0007,
828
+ "loss": 2.142623710632324,
829
+ "step": 2340
830
+ },
831
+ {
832
+ "epoch": 0.15908324907313784,
833
+ "grad_norm": 0.333984375,
834
+ "learning_rate": 0.0007,
835
+ "loss": 2.1457294464111327,
836
+ "step": 2360
837
+ },
838
+ {
839
+ "epoch": 0.1604314122008763,
840
+ "grad_norm": 0.33203125,
841
+ "learning_rate": 0.0007,
842
+ "loss": 2.1467227935791016,
843
+ "step": 2380
844
+ },
845
+ {
846
+ "epoch": 0.16177957532861476,
847
+ "grad_norm": 0.369140625,
848
+ "learning_rate": 0.0007,
849
+ "loss": 2.1286346435546877,
850
+ "step": 2400
851
+ }
852
+ ],
853
+ "logging_steps": 20,
854
+ "max_steps": 2500,
855
+ "num_input_tokens_seen": 0,
856
+ "num_train_epochs": 1,
857
+ "save_steps": 100,
858
+ "stateful_callbacks": {
859
+ "TrainerControl": {
860
+ "args": {
861
+ "should_epoch_stop": false,
862
+ "should_evaluate": false,
863
+ "should_log": false,
864
+ "should_save": true,
865
+ "should_training_stop": false
866
+ },
867
+ "attributes": {}
868
+ }
869
+ },
870
+ "total_flos": 7743081126297600.0,
871
+ "train_batch_size": 64,
872
+ "trial_name": null,
873
+ "trial_params": null
874
+ }
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2400/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:efb8cca2a9ea6ff541b7562ec4e2e31a24b5072ef5136a6523ae9f9c0012caa0
3
+ size 4920
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "situglu_low",
3
+ "architectures": [
4
+ "TinyLlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 2,
11
+ "head_dim": 32,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 128,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 256,
16
+ "max_position_embeddings": 512,
17
+ "mlp_bias": false,
18
+ "mlp_type": "glu",
19
+ "model_type": "tiny_llama",
20
+ "num_attention_heads": 4,
21
+ "num_hidden_layers": 100,
22
+ "num_key_value_heads": 4,
23
+ "pad_token_id": 0,
24
+ "pretraining_tp": 1,
25
+ "rms_norm_eps": 1e-06,
26
+ "rope_parameters": {
27
+ "rope_theta": 10000.0,
28
+ "rope_type": "default"
29
+ },
30
+ "tie_word_embeddings": true,
31
+ "tokenizer_name": "w-ahmad/tiny-stories-tokenizer",
32
+ "transformers_version": "5.16.0.dev0",
33
+ "use_cache": false,
34
+ "vocab_size": 4096,
35
+ "waleed_beta": 10.0
36
+ }
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3e7b15542270620c3fcc10261277bc54e2e41ceaa20b3386de734d4185481f8e
3
+ size 33967272
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0023a612ec368fd1cfcbd9835329e9dae7dd366352314799cc790ef88358f8fe
3
+ size 68504996
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3cf9097d4513154245c48236b6ec5137b7ee2a21c9f58f2cba798ea275c6026f
3
+ size 14244
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fae3f7a9f86d9663c06f6e22221bc9e6879cd9691e2b1bba04ee6412e2572f45
3
+ size 1064
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/tokenizer_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|endoftext|>",
5
+ "eos_token": "<|endoftext|>",
6
+ "errors": "replace",
7
+ "is_local": false,
8
+ "local_files_only": false,
9
+ "model_max_length": 1024,
10
+ "pad_token": "<|endoftext|>",
11
+ "tokenizer_class": "GPT2Tokenizer",
12
+ "unk_token": "<|endoftext|>"
13
+ }
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/trainer_state.json ADDED
@@ -0,0 +1,925 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.16852039096730703,
6
+ "eval_steps": 2498,
7
+ "global_step": 2500,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0013481631277384564,
14
+ "grad_norm": 2.078125,
15
+ "learning_rate": 2.66e-05,
16
+ "loss": 8.299667358398438,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.3515625,
22
+ "learning_rate": 5.46e-05,
23
+ "loss": 8.047861480712891,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 1.2734375,
29
+ "learning_rate": 8.259999999999999e-05,
30
+ "loss": 7.838032531738281,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 1.265625,
36
+ "learning_rate": 0.0001106,
37
+ "loss": 7.572964477539062,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 1.1953125,
43
+ "learning_rate": 0.0001386,
44
+ "loss": 7.233274841308594,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 1.1328125,
50
+ "learning_rate": 0.00016659999999999998,
51
+ "loss": 6.860704803466797,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 0.94921875,
57
+ "learning_rate": 0.00019460000000000001,
58
+ "loss": 6.517856597900391,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 0.8046875,
64
+ "learning_rate": 0.0002226,
65
+ "loss": 6.239397811889648,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 0.65234375,
71
+ "learning_rate": 0.00025059999999999997,
72
+ "loss": 6.039968109130859,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 0.5625,
78
+ "learning_rate": 0.0002786,
79
+ "loss": 5.863864517211914,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.01482979440512302,
84
+ "grad_norm": 0.67578125,
85
+ "learning_rate": 0.00030659999999999997,
86
+ "loss": 5.679952239990234,
87
+ "step": 220
88
+ },
89
+ {
90
+ "epoch": 0.016177957532861477,
91
+ "grad_norm": 1.484375,
92
+ "learning_rate": 0.0003346,
93
+ "loss": 5.470568084716797,
94
+ "step": 240
95
+ },
96
+ {
97
+ "epoch": 0.01752612066059993,
98
+ "grad_norm": 1.9765625,
99
+ "learning_rate": 0.00036260000000000003,
100
+ "loss": 5.266762542724609,
101
+ "step": 260
102
+ },
103
+ {
104
+ "epoch": 0.01887428378833839,
105
+ "grad_norm": 0.6015625,
106
+ "learning_rate": 0.0003906,
107
+ "loss": 5.04127197265625,
108
+ "step": 280
109
+ },
110
+ {
111
+ "epoch": 0.020222446916076844,
112
+ "grad_norm": 1.0859375,
113
+ "learning_rate": 0.0004186,
114
+ "loss": 4.803184509277344,
115
+ "step": 300
116
+ },
117
+ {
118
+ "epoch": 0.021570610043815303,
119
+ "grad_norm": 1.15625,
120
+ "learning_rate": 0.0004466,
121
+ "loss": 4.589862823486328,
122
+ "step": 320
123
+ },
124
+ {
125
+ "epoch": 0.022918773171553757,
126
+ "grad_norm": 0.6953125,
127
+ "learning_rate": 0.00047460000000000004,
128
+ "loss": 4.416217803955078,
129
+ "step": 340
130
+ },
131
+ {
132
+ "epoch": 0.024266936299292215,
133
+ "grad_norm": 0.8046875,
134
+ "learning_rate": 0.0005026,
135
+ "loss": 4.270945739746094,
136
+ "step": 360
137
+ },
138
+ {
139
+ "epoch": 0.02561509942703067,
140
+ "grad_norm": 0.75,
141
+ "learning_rate": 0.0005306,
142
+ "loss": 4.125864028930664,
143
+ "step": 380
144
+ },
145
+ {
146
+ "epoch": 0.026963262554769128,
147
+ "grad_norm": 0.5625,
148
+ "learning_rate": 0.0005586,
149
+ "loss": 3.992333984375,
150
+ "step": 400
151
+ },
152
+ {
153
+ "epoch": 0.028311425682507583,
154
+ "grad_norm": 0.53515625,
155
+ "learning_rate": 0.0005866,
156
+ "loss": 3.8864391326904295,
157
+ "step": 420
158
+ },
159
+ {
160
+ "epoch": 0.02965958881024604,
161
+ "grad_norm": 0.86328125,
162
+ "learning_rate": 0.0006146,
163
+ "loss": 3.819641876220703,
164
+ "step": 440
165
+ },
166
+ {
167
+ "epoch": 0.031007751937984496,
168
+ "grad_norm": 0.7109375,
169
+ "learning_rate": 0.0006426,
170
+ "loss": 3.7153347015380858,
171
+ "step": 460
172
+ },
173
+ {
174
+ "epoch": 0.032355915065722954,
175
+ "grad_norm": 0.53515625,
176
+ "learning_rate": 0.0006705999999999999,
177
+ "loss": 3.6391063690185548,
178
+ "step": 480
179
+ },
180
+ {
181
+ "epoch": 0.03370407819346141,
182
+ "grad_norm": 0.59765625,
183
+ "learning_rate": 0.0006986,
184
+ "loss": 3.5594459533691407,
185
+ "step": 500
186
+ },
187
+ {
188
+ "epoch": 0.03505224132119986,
189
+ "grad_norm": 0.4296875,
190
+ "learning_rate": 0.0007,
191
+ "loss": 3.493515396118164,
192
+ "step": 520
193
+ },
194
+ {
195
+ "epoch": 0.03640040444893832,
196
+ "grad_norm": 0.640625,
197
+ "learning_rate": 0.0007,
198
+ "loss": 3.41654052734375,
199
+ "step": 540
200
+ },
201
+ {
202
+ "epoch": 0.03774856757667678,
203
+ "grad_norm": 0.484375,
204
+ "learning_rate": 0.0007,
205
+ "loss": 3.366413879394531,
206
+ "step": 560
207
+ },
208
+ {
209
+ "epoch": 0.03909673070441524,
210
+ "grad_norm": 0.423828125,
211
+ "learning_rate": 0.0007,
212
+ "loss": 3.3078380584716798,
213
+ "step": 580
214
+ },
215
+ {
216
+ "epoch": 0.04044489383215369,
217
+ "grad_norm": 0.423828125,
218
+ "learning_rate": 0.0007,
219
+ "loss": 3.2723438262939455,
220
+ "step": 600
221
+ },
222
+ {
223
+ "epoch": 0.04179305695989215,
224
+ "grad_norm": 0.431640625,
225
+ "learning_rate": 0.0007,
226
+ "loss": 3.2081642150878906,
227
+ "step": 620
228
+ },
229
+ {
230
+ "epoch": 0.043141220087630605,
231
+ "grad_norm": 0.44921875,
232
+ "learning_rate": 0.0007,
233
+ "loss": 3.1861698150634767,
234
+ "step": 640
235
+ },
236
+ {
237
+ "epoch": 0.044489383215369056,
238
+ "grad_norm": 0.40625,
239
+ "learning_rate": 0.0007,
240
+ "loss": 3.126705360412598,
241
+ "step": 660
242
+ },
243
+ {
244
+ "epoch": 0.045837546343107514,
245
+ "grad_norm": 0.423828125,
246
+ "learning_rate": 0.0007,
247
+ "loss": 3.1009654998779297,
248
+ "step": 680
249
+ },
250
+ {
251
+ "epoch": 0.04718570947084597,
252
+ "grad_norm": 0.41015625,
253
+ "learning_rate": 0.0007,
254
+ "loss": 3.0569204330444335,
255
+ "step": 700
256
+ },
257
+ {
258
+ "epoch": 0.04853387259858443,
259
+ "grad_norm": 0.390625,
260
+ "learning_rate": 0.0007,
261
+ "loss": 3.0219058990478516,
262
+ "step": 720
263
+ },
264
+ {
265
+ "epoch": 0.04988203572632288,
266
+ "grad_norm": 0.447265625,
267
+ "learning_rate": 0.0007,
268
+ "loss": 2.9834646224975585,
269
+ "step": 740
270
+ },
271
+ {
272
+ "epoch": 0.05123019885406134,
273
+ "grad_norm": 0.39453125,
274
+ "learning_rate": 0.0007,
275
+ "loss": 2.9665781021118165,
276
+ "step": 760
277
+ },
278
+ {
279
+ "epoch": 0.0525783619817998,
280
+ "grad_norm": 0.43359375,
281
+ "learning_rate": 0.0007,
282
+ "loss": 2.9737390518188476,
283
+ "step": 780
284
+ },
285
+ {
286
+ "epoch": 0.053926525109538256,
287
+ "grad_norm": 0.41015625,
288
+ "learning_rate": 0.0007,
289
+ "loss": 2.9195796966552736,
290
+ "step": 800
291
+ },
292
+ {
293
+ "epoch": 0.05527468823727671,
294
+ "grad_norm": 0.41015625,
295
+ "learning_rate": 0.0007,
296
+ "loss": 2.9032770156860352,
297
+ "step": 820
298
+ },
299
+ {
300
+ "epoch": 0.056622851365015166,
301
+ "grad_norm": 0.4296875,
302
+ "learning_rate": 0.0007,
303
+ "loss": 2.8703426361083983,
304
+ "step": 840
305
+ },
306
+ {
307
+ "epoch": 0.057971014492753624,
308
+ "grad_norm": 0.384765625,
309
+ "learning_rate": 0.0007,
310
+ "loss": 2.8253679275512695,
311
+ "step": 860
312
+ },
313
+ {
314
+ "epoch": 0.05931917762049208,
315
+ "grad_norm": 0.390625,
316
+ "learning_rate": 0.0007,
317
+ "loss": 2.802602195739746,
318
+ "step": 880
319
+ },
320
+ {
321
+ "epoch": 0.06066734074823053,
322
+ "grad_norm": 0.40625,
323
+ "learning_rate": 0.0007,
324
+ "loss": 2.7884506225585937,
325
+ "step": 900
326
+ },
327
+ {
328
+ "epoch": 0.06201550387596899,
329
+ "grad_norm": 0.3828125,
330
+ "learning_rate": 0.0007,
331
+ "loss": 2.7588369369506838,
332
+ "step": 920
333
+ },
334
+ {
335
+ "epoch": 0.06336366700370745,
336
+ "grad_norm": 0.41796875,
337
+ "learning_rate": 0.0007,
338
+ "loss": 2.7435585021972657,
339
+ "step": 940
340
+ },
341
+ {
342
+ "epoch": 0.06471183013144591,
343
+ "grad_norm": 0.44921875,
344
+ "learning_rate": 0.0007,
345
+ "loss": 2.717236328125,
346
+ "step": 960
347
+ },
348
+ {
349
+ "epoch": 0.06605999325918437,
350
+ "grad_norm": 0.373046875,
351
+ "learning_rate": 0.0007,
352
+ "loss": 2.699039268493652,
353
+ "step": 980
354
+ },
355
+ {
356
+ "epoch": 0.06740815638692282,
357
+ "grad_norm": 0.41015625,
358
+ "learning_rate": 0.0007,
359
+ "loss": 2.697044563293457,
360
+ "step": 1000
361
+ },
362
+ {
363
+ "epoch": 0.06875631951466127,
364
+ "grad_norm": 0.361328125,
365
+ "learning_rate": 0.0007,
366
+ "loss": 2.6641401290893554,
367
+ "step": 1020
368
+ },
369
+ {
370
+ "epoch": 0.07010448264239973,
371
+ "grad_norm": 0.376953125,
372
+ "learning_rate": 0.0007,
373
+ "loss": 2.6494163513183593,
374
+ "step": 1040
375
+ },
376
+ {
377
+ "epoch": 0.07145264577013818,
378
+ "grad_norm": 0.416015625,
379
+ "learning_rate": 0.0007,
380
+ "loss": 2.631260299682617,
381
+ "step": 1060
382
+ },
383
+ {
384
+ "epoch": 0.07280080889787664,
385
+ "grad_norm": 0.37890625,
386
+ "learning_rate": 0.0007,
387
+ "loss": 2.612077331542969,
388
+ "step": 1080
389
+ },
390
+ {
391
+ "epoch": 0.0741489720256151,
392
+ "grad_norm": 0.40234375,
393
+ "learning_rate": 0.0007,
394
+ "loss": 2.606185722351074,
395
+ "step": 1100
396
+ },
397
+ {
398
+ "epoch": 0.07549713515335356,
399
+ "grad_norm": 0.388671875,
400
+ "learning_rate": 0.0007,
401
+ "loss": 2.570285606384277,
402
+ "step": 1120
403
+ },
404
+ {
405
+ "epoch": 0.07684529828109202,
406
+ "grad_norm": 0.41796875,
407
+ "learning_rate": 0.0007,
408
+ "loss": 2.557082176208496,
409
+ "step": 1140
410
+ },
411
+ {
412
+ "epoch": 0.07819346140883048,
413
+ "grad_norm": 0.37890625,
414
+ "learning_rate": 0.0007,
415
+ "loss": 2.5526586532592774,
416
+ "step": 1160
417
+ },
418
+ {
419
+ "epoch": 0.07954162453656892,
420
+ "grad_norm": 0.388671875,
421
+ "learning_rate": 0.0007,
422
+ "loss": 2.547346305847168,
423
+ "step": 1180
424
+ },
425
+ {
426
+ "epoch": 0.08088978766430738,
427
+ "grad_norm": 0.3828125,
428
+ "learning_rate": 0.0007,
429
+ "loss": 2.5366180419921873,
430
+ "step": 1200
431
+ },
432
+ {
433
+ "epoch": 0.08223795079204584,
434
+ "grad_norm": 0.359375,
435
+ "learning_rate": 0.0007,
436
+ "loss": 2.5206745147705076,
437
+ "step": 1220
438
+ },
439
+ {
440
+ "epoch": 0.0835861139197843,
441
+ "grad_norm": 0.36328125,
442
+ "learning_rate": 0.0007,
443
+ "loss": 2.506051254272461,
444
+ "step": 1240
445
+ },
446
+ {
447
+ "epoch": 0.08493427704752275,
448
+ "grad_norm": 0.376953125,
449
+ "learning_rate": 0.0007,
450
+ "loss": 2.494335174560547,
451
+ "step": 1260
452
+ },
453
+ {
454
+ "epoch": 0.08628244017526121,
455
+ "grad_norm": 0.384765625,
456
+ "learning_rate": 0.0007,
457
+ "loss": 2.4746730804443358,
458
+ "step": 1280
459
+ },
460
+ {
461
+ "epoch": 0.08763060330299967,
462
+ "grad_norm": 0.384765625,
463
+ "learning_rate": 0.0007,
464
+ "loss": 2.4673040390014647,
465
+ "step": 1300
466
+ },
467
+ {
468
+ "epoch": 0.08897876643073811,
469
+ "grad_norm": 0.359375,
470
+ "learning_rate": 0.0007,
471
+ "loss": 2.458811950683594,
472
+ "step": 1320
473
+ },
474
+ {
475
+ "epoch": 0.09032692955847657,
476
+ "grad_norm": 0.40234375,
477
+ "learning_rate": 0.0007,
478
+ "loss": 2.444304275512695,
479
+ "step": 1340
480
+ },
481
+ {
482
+ "epoch": 0.09167509268621503,
483
+ "grad_norm": 0.3984375,
484
+ "learning_rate": 0.0007,
485
+ "loss": 2.4439945220947266,
486
+ "step": 1360
487
+ },
488
+ {
489
+ "epoch": 0.09302325581395349,
490
+ "grad_norm": 0.341796875,
491
+ "learning_rate": 0.0007,
492
+ "loss": 2.4202173233032225,
493
+ "step": 1380
494
+ },
495
+ {
496
+ "epoch": 0.09437141894169195,
497
+ "grad_norm": 0.36328125,
498
+ "learning_rate": 0.0007,
499
+ "loss": 2.4103469848632812,
500
+ "step": 1400
501
+ },
502
+ {
503
+ "epoch": 0.0957195820694304,
504
+ "grad_norm": 0.39453125,
505
+ "learning_rate": 0.0007,
506
+ "loss": 2.4228416442871095,
507
+ "step": 1420
508
+ },
509
+ {
510
+ "epoch": 0.09706774519716886,
511
+ "grad_norm": 0.3515625,
512
+ "learning_rate": 0.0007,
513
+ "loss": 2.3987619400024416,
514
+ "step": 1440
515
+ },
516
+ {
517
+ "epoch": 0.09841590832490732,
518
+ "grad_norm": 0.38671875,
519
+ "learning_rate": 0.0007,
520
+ "loss": 2.3866939544677734,
521
+ "step": 1460
522
+ },
523
+ {
524
+ "epoch": 0.09976407145264576,
525
+ "grad_norm": 0.359375,
526
+ "learning_rate": 0.0007,
527
+ "loss": 2.3730697631835938,
528
+ "step": 1480
529
+ },
530
+ {
531
+ "epoch": 0.10111223458038422,
532
+ "grad_norm": 0.359375,
533
+ "learning_rate": 0.0007,
534
+ "loss": 2.3593896865844726,
535
+ "step": 1500
536
+ },
537
+ {
538
+ "epoch": 0.10246039770812268,
539
+ "grad_norm": 0.37890625,
540
+ "learning_rate": 0.0007,
541
+ "loss": 2.362796401977539,
542
+ "step": 1520
543
+ },
544
+ {
545
+ "epoch": 0.10380856083586114,
546
+ "grad_norm": 0.3828125,
547
+ "learning_rate": 0.0007,
548
+ "loss": 2.3507095336914063,
549
+ "step": 1540
550
+ },
551
+ {
552
+ "epoch": 0.1051567239635996,
553
+ "grad_norm": 0.37890625,
554
+ "learning_rate": 0.0007,
555
+ "loss": 2.339781951904297,
556
+ "step": 1560
557
+ },
558
+ {
559
+ "epoch": 0.10650488709133805,
560
+ "grad_norm": 0.353515625,
561
+ "learning_rate": 0.0007,
562
+ "loss": 2.356917953491211,
563
+ "step": 1580
564
+ },
565
+ {
566
+ "epoch": 0.10785305021907651,
567
+ "grad_norm": 0.392578125,
568
+ "learning_rate": 0.0007,
569
+ "loss": 2.325514030456543,
570
+ "step": 1600
571
+ },
572
+ {
573
+ "epoch": 0.10920121334681497,
574
+ "grad_norm": 0.33984375,
575
+ "learning_rate": 0.0007,
576
+ "loss": 2.3245080947875976,
577
+ "step": 1620
578
+ },
579
+ {
580
+ "epoch": 0.11054937647455342,
581
+ "grad_norm": 0.349609375,
582
+ "learning_rate": 0.0007,
583
+ "loss": 2.3359960556030273,
584
+ "step": 1640
585
+ },
586
+ {
587
+ "epoch": 0.11189753960229187,
588
+ "grad_norm": 0.3359375,
589
+ "learning_rate": 0.0007,
590
+ "loss": 2.3171085357666015,
591
+ "step": 1660
592
+ },
593
+ {
594
+ "epoch": 0.11324570273003033,
595
+ "grad_norm": 0.37890625,
596
+ "learning_rate": 0.0007,
597
+ "loss": 2.299543571472168,
598
+ "step": 1680
599
+ },
600
+ {
601
+ "epoch": 0.11459386585776879,
602
+ "grad_norm": 0.3515625,
603
+ "learning_rate": 0.0007,
604
+ "loss": 2.309956932067871,
605
+ "step": 1700
606
+ },
607
+ {
608
+ "epoch": 0.11594202898550725,
609
+ "grad_norm": 0.3828125,
610
+ "learning_rate": 0.0007,
611
+ "loss": 2.2854846954345702,
612
+ "step": 1720
613
+ },
614
+ {
615
+ "epoch": 0.1172901921132457,
616
+ "grad_norm": 0.349609375,
617
+ "learning_rate": 0.0007,
618
+ "loss": 2.2760204315185546,
619
+ "step": 1740
620
+ },
621
+ {
622
+ "epoch": 0.11863835524098416,
623
+ "grad_norm": 0.3515625,
624
+ "learning_rate": 0.0007,
625
+ "loss": 2.279789924621582,
626
+ "step": 1760
627
+ },
628
+ {
629
+ "epoch": 0.11998651836872262,
630
+ "grad_norm": 0.3671875,
631
+ "learning_rate": 0.0007,
632
+ "loss": 2.2942825317382813,
633
+ "step": 1780
634
+ },
635
+ {
636
+ "epoch": 0.12133468149646107,
637
+ "grad_norm": 0.341796875,
638
+ "learning_rate": 0.0007,
639
+ "loss": 2.2891586303710936,
640
+ "step": 1800
641
+ },
642
+ {
643
+ "epoch": 0.12268284462419952,
644
+ "grad_norm": 0.3828125,
645
+ "learning_rate": 0.0007,
646
+ "loss": 2.279863166809082,
647
+ "step": 1820
648
+ },
649
+ {
650
+ "epoch": 0.12403100775193798,
651
+ "grad_norm": 0.359375,
652
+ "learning_rate": 0.0007,
653
+ "loss": 2.2473472595214843,
654
+ "step": 1840
655
+ },
656
+ {
657
+ "epoch": 0.12537917087967643,
658
+ "grad_norm": 0.33984375,
659
+ "learning_rate": 0.0007,
660
+ "loss": 2.248352813720703,
661
+ "step": 1860
662
+ },
663
+ {
664
+ "epoch": 0.1267273340074149,
665
+ "grad_norm": 0.322265625,
666
+ "learning_rate": 0.0007,
667
+ "loss": 2.239566612243652,
668
+ "step": 1880
669
+ },
670
+ {
671
+ "epoch": 0.12807549713515334,
672
+ "grad_norm": 0.3359375,
673
+ "learning_rate": 0.0007,
674
+ "loss": 2.2460857391357423,
675
+ "step": 1900
676
+ },
677
+ {
678
+ "epoch": 0.12942366026289182,
679
+ "grad_norm": 0.345703125,
680
+ "learning_rate": 0.0007,
681
+ "loss": 2.222591209411621,
682
+ "step": 1920
683
+ },
684
+ {
685
+ "epoch": 0.13077182339063026,
686
+ "grad_norm": 0.35546875,
687
+ "learning_rate": 0.0007,
688
+ "loss": 2.2339385986328124,
689
+ "step": 1940
690
+ },
691
+ {
692
+ "epoch": 0.13211998651836873,
693
+ "grad_norm": 0.337890625,
694
+ "learning_rate": 0.0007,
695
+ "loss": 2.2090906143188476,
696
+ "step": 1960
697
+ },
698
+ {
699
+ "epoch": 0.13346814964610718,
700
+ "grad_norm": 0.36328125,
701
+ "learning_rate": 0.0007,
702
+ "loss": 2.2296125411987306,
703
+ "step": 1980
704
+ },
705
+ {
706
+ "epoch": 0.13481631277384565,
707
+ "grad_norm": 0.345703125,
708
+ "learning_rate": 0.0007,
709
+ "loss": 2.20355110168457,
710
+ "step": 2000
711
+ },
712
+ {
713
+ "epoch": 0.1361644759015841,
714
+ "grad_norm": 0.34765625,
715
+ "learning_rate": 0.0007,
716
+ "loss": 2.211279296875,
717
+ "step": 2020
718
+ },
719
+ {
720
+ "epoch": 0.13751263902932254,
721
+ "grad_norm": 0.37890625,
722
+ "learning_rate": 0.0007,
723
+ "loss": 2.1930706024169924,
724
+ "step": 2040
725
+ },
726
+ {
727
+ "epoch": 0.138860802157061,
728
+ "grad_norm": 0.3359375,
729
+ "learning_rate": 0.0007,
730
+ "loss": 2.211184501647949,
731
+ "step": 2060
732
+ },
733
+ {
734
+ "epoch": 0.14020896528479945,
735
+ "grad_norm": 0.333984375,
736
+ "learning_rate": 0.0007,
737
+ "loss": 2.1871965408325194,
738
+ "step": 2080
739
+ },
740
+ {
741
+ "epoch": 0.14155712841253792,
742
+ "grad_norm": 0.33203125,
743
+ "learning_rate": 0.0007,
744
+ "loss": 2.178455352783203,
745
+ "step": 2100
746
+ },
747
+ {
748
+ "epoch": 0.14290529154027637,
749
+ "grad_norm": 0.35546875,
750
+ "learning_rate": 0.0007,
751
+ "loss": 2.186496925354004,
752
+ "step": 2120
753
+ },
754
+ {
755
+ "epoch": 0.14425345466801484,
756
+ "grad_norm": 0.341796875,
757
+ "learning_rate": 0.0007,
758
+ "loss": 2.1872365951538084,
759
+ "step": 2140
760
+ },
761
+ {
762
+ "epoch": 0.14560161779575329,
763
+ "grad_norm": 0.337890625,
764
+ "learning_rate": 0.0007,
765
+ "loss": 2.1733598709106445,
766
+ "step": 2160
767
+ },
768
+ {
769
+ "epoch": 0.14694978092349173,
770
+ "grad_norm": 0.3359375,
771
+ "learning_rate": 0.0007,
772
+ "loss": 2.1814584732055664,
773
+ "step": 2180
774
+ },
775
+ {
776
+ "epoch": 0.1482979440512302,
777
+ "grad_norm": 0.328125,
778
+ "learning_rate": 0.0007,
779
+ "loss": 2.157553291320801,
780
+ "step": 2200
781
+ },
782
+ {
783
+ "epoch": 0.14964610717896865,
784
+ "grad_norm": 0.341796875,
785
+ "learning_rate": 0.0007,
786
+ "loss": 2.1652835845947265,
787
+ "step": 2220
788
+ },
789
+ {
790
+ "epoch": 0.15099427030670712,
791
+ "grad_norm": 0.34375,
792
+ "learning_rate": 0.0007,
793
+ "loss": 2.1730274200439452,
794
+ "step": 2240
795
+ },
796
+ {
797
+ "epoch": 0.15234243343444556,
798
+ "grad_norm": 0.353515625,
799
+ "learning_rate": 0.0007,
800
+ "loss": 2.1498680114746094,
801
+ "step": 2260
802
+ },
803
+ {
804
+ "epoch": 0.15369059656218403,
805
+ "grad_norm": 0.326171875,
806
+ "learning_rate": 0.0007,
807
+ "loss": 2.1618186950683596,
808
+ "step": 2280
809
+ },
810
+ {
811
+ "epoch": 0.15503875968992248,
812
+ "grad_norm": 0.337890625,
813
+ "learning_rate": 0.0007,
814
+ "loss": 2.142350959777832,
815
+ "step": 2300
816
+ },
817
+ {
818
+ "epoch": 0.15638692281766095,
819
+ "grad_norm": 0.333984375,
820
+ "learning_rate": 0.0007,
821
+ "loss": 2.1518985748291017,
822
+ "step": 2320
823
+ },
824
+ {
825
+ "epoch": 0.1577350859453994,
826
+ "grad_norm": 0.337890625,
827
+ "learning_rate": 0.0007,
828
+ "loss": 2.142623710632324,
829
+ "step": 2340
830
+ },
831
+ {
832
+ "epoch": 0.15908324907313784,
833
+ "grad_norm": 0.333984375,
834
+ "learning_rate": 0.0007,
835
+ "loss": 2.1457294464111327,
836
+ "step": 2360
837
+ },
838
+ {
839
+ "epoch": 0.1604314122008763,
840
+ "grad_norm": 0.33203125,
841
+ "learning_rate": 0.0007,
842
+ "loss": 2.1467227935791016,
843
+ "step": 2380
844
+ },
845
+ {
846
+ "epoch": 0.16177957532861476,
847
+ "grad_norm": 0.369140625,
848
+ "learning_rate": 0.0007,
849
+ "loss": 2.1286346435546877,
850
+ "step": 2400
851
+ },
852
+ {
853
+ "epoch": 0.16312773845635323,
854
+ "grad_norm": 0.34375,
855
+ "learning_rate": 0.0007,
856
+ "loss": 2.1322181701660154,
857
+ "step": 2420
858
+ },
859
+ {
860
+ "epoch": 0.16447590158409167,
861
+ "grad_norm": 0.330078125,
862
+ "learning_rate": 0.0007,
863
+ "loss": 2.119257354736328,
864
+ "step": 2440
865
+ },
866
+ {
867
+ "epoch": 0.16582406471183014,
868
+ "grad_norm": 0.341796875,
869
+ "learning_rate": 0.0007,
870
+ "loss": 2.1314579010009767,
871
+ "step": 2460
872
+ },
873
+ {
874
+ "epoch": 0.1671722278395686,
875
+ "grad_norm": 0.34765625,
876
+ "learning_rate": 0.0007,
877
+ "loss": 2.1324623107910154,
878
+ "step": 2480
879
+ },
880
+ {
881
+ "epoch": 0.1683855746545332,
882
+ "eval_loss": 2.125929832458496,
883
+ "eval_runtime": 16.8275,
884
+ "eval_samples_per_second": 566.157,
885
+ "eval_steps_per_second": 0.594,
886
+ "step": 2498
887
+ },
888
+ {
889
+ "epoch": 0.16852039096730703,
890
+ "grad_norm": 0.333984375,
891
+ "learning_rate": 0.0007,
892
+ "loss": 2.1140785217285156,
893
+ "step": 2500
894
+ },
895
+ {
896
+ "epoch": 0.16852039096730703,
897
+ "eval_loss": 2.1262547969818115,
898
+ "eval_runtime": 16.635,
899
+ "eval_samples_per_second": 572.71,
900
+ "eval_steps_per_second": 0.601,
901
+ "step": 2500
902
+ }
903
+ ],
904
+ "logging_steps": 20,
905
+ "max_steps": 2500,
906
+ "num_input_tokens_seen": 0,
907
+ "num_train_epochs": 1,
908
+ "save_steps": 100,
909
+ "stateful_callbacks": {
910
+ "TrainerControl": {
911
+ "args": {
912
+ "should_epoch_stop": false,
913
+ "should_evaluate": false,
914
+ "should_log": false,
915
+ "should_save": true,
916
+ "should_training_stop": true
917
+ },
918
+ "attributes": {}
919
+ }
920
+ },
921
+ "total_flos": 8065709506560000.0,
922
+ "train_batch_size": 64,
923
+ "trial_name": null,
924
+ "trial_params": null
925
+ }
zain/Activation/out/glu-situglu_low-100L_run/checkpoint-2500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:efb8cca2a9ea6ff541b7562ec4e2e31a24b5072ef5136a6523ae9f9c0012caa0
3
+ size 4920
zain/Activation/out/glu-situglu_low-100L_run/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "situglu_low",
3
+ "architectures": [
4
+ "TinyLlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 2,
11
+ "head_dim": 32,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 128,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 256,
16
+ "max_position_embeddings": 512,
17
+ "mlp_bias": false,
18
+ "mlp_type": "glu",
19
+ "model_type": "tiny_llama",
20
+ "num_attention_heads": 4,
21
+ "num_hidden_layers": 100,
22
+ "num_key_value_heads": 4,
23
+ "pad_token_id": 0,
24
+ "pretraining_tp": 1,
25
+ "rms_norm_eps": 1e-06,
26
+ "rope_parameters": {
27
+ "rope_theta": 10000.0,
28
+ "rope_type": "default"
29
+ },
30
+ "tie_word_embeddings": true,
31
+ "tokenizer_name": "w-ahmad/tiny-stories-tokenizer",
32
+ "transformers_version": "5.16.0.dev0",
33
+ "use_cache": false,
34
+ "vocab_size": 4096,
35
+ "waleed_beta": 10.0
36
+ }
zain/Activation/out/glu-situglu_low-100L_run/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3e7b15542270620c3fcc10261277bc54e2e41ceaa20b3386de734d4185481f8e
3
+ size 33967272
zain/Activation/out/glu-situglu_low-100L_run/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-situglu_low-100L_run/tokenizer_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|endoftext|>",
5
+ "eos_token": "<|endoftext|>",
6
+ "errors": "replace",
7
+ "is_local": false,
8
+ "local_files_only": false,
9
+ "model_max_length": 1024,
10
+ "pad_token": "<|endoftext|>",
11
+ "tokenizer_class": "GPT2Tokenizer",
12
+ "unk_token": "<|endoftext|>"
13
+ }
zain/Activation/out/glu-situglu_low-100L_run/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:efb8cca2a9ea6ff541b7562ec4e2e31a24b5072ef5136a6523ae9f9c0012caa0
3
+ size 4920
zain/Activation/out/glu-situglu_low-100L_run/training_log.jsonl CHANGED
@@ -109,3 +109,21 @@
109
  {"step": 2180, "epoch": 0.14694978092349173, "timestamp": 1786576852.170506, "loss": 2.1814584732055664, "grad_norm": 0.3359375, "learning_rate": 0.0007, "train/total_time_seconds": 579.0546316727996, "train/time_per_step_avg": 0.2629444441944361, "train/epoch_time_elapsed": 696.1408161707222, "train/estimated_remaining_minutes": 1.4166474169365129}
110
  {"step": 2200, "epoch": 0.1482979440512302, "timestamp": 1786576858.479852, "loss": 2.157553291320801, "grad_norm": 0.328125, "learning_rate": 0.0007, "train/total_time_seconds": 584.3338468708098, "train/time_per_step_avg": 0.2632886817678809, "train/epoch_time_elapsed": 702.4501621201634, "train/estimated_remaining_minutes": 1.3280314701609313}
111
  {"step": 2220, "epoch": 0.14964610717896865, "timestamp": 1786576865.4107788, "loss": 2.1652835845947265, "grad_norm": 0.341796875, "learning_rate": 0.0007, "train/total_time_seconds": 589.6443728506565, "train/time_per_step_avg": 0.26375446785241363, "train/epoch_time_elapsed": 709.3810888081789, "train/estimated_remaining_minutes": 1.2394926756620408}
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
109
  {"step": 2180, "epoch": 0.14694978092349173, "timestamp": 1786576852.170506, "loss": 2.1814584732055664, "grad_norm": 0.3359375, "learning_rate": 0.0007, "train/total_time_seconds": 579.0546316727996, "train/time_per_step_avg": 0.2629444441944361, "train/epoch_time_elapsed": 696.1408161707222, "train/estimated_remaining_minutes": 1.4166474169365129}
110
  {"step": 2200, "epoch": 0.1482979440512302, "timestamp": 1786576858.479852, "loss": 2.157553291320801, "grad_norm": 0.328125, "learning_rate": 0.0007, "train/total_time_seconds": 584.3338468708098, "train/time_per_step_avg": 0.2632886817678809, "train/epoch_time_elapsed": 702.4501621201634, "train/estimated_remaining_minutes": 1.3280314701609313}
111
  {"step": 2220, "epoch": 0.14964610717896865, "timestamp": 1786576865.4107788, "loss": 2.1652835845947265, "grad_norm": 0.341796875, "learning_rate": 0.0007, "train/total_time_seconds": 589.6443728506565, "train/time_per_step_avg": 0.26375446785241363, "train/epoch_time_elapsed": 709.3810888081789, "train/estimated_remaining_minutes": 1.2394926756620408}
112
+ {"step": 2240, "epoch": 0.15099427030670712, "timestamp": 1786576871.690686, "loss": 2.1730274200439452, "grad_norm": 0.34375, "learning_rate": 0.0007, "train/total_time_seconds": 594.9144274070859, "train/time_per_step_avg": 0.2637659421563148, "train/epoch_time_elapsed": 715.6609964966774, "train/estimated_remaining_minutes": 1.1508761244482315}
113
+ {"step": 2260, "epoch": 0.15234243343444556, "timestamp": 1786576878.035702, "loss": 2.1498680114746094, "grad_norm": 0.353515625, "learning_rate": 0.0007, "train/total_time_seconds": 600.2152769267559, "train/time_per_step_avg": 0.26393290534615516, "train/epoch_time_elapsed": 722.0060120299459, "train/estimated_remaining_minutes": 1.0623279237641698}
114
+ {"step": 2280, "epoch": 0.15369059656218403, "timestamp": 1786576884.2997983, "loss": 2.1618186950683596, "grad_norm": 0.326171875, "learning_rate": 0.0007, "train/total_time_seconds": 605.477860070765, "train/time_per_step_avg": 0.2642322839796543, "train/epoch_time_elapsed": 728.2701090201735, "train/estimated_remaining_minutes": 0.9737217047921659}
115
+ {"step": 2300, "epoch": 0.15503875968992248, "timestamp": 1786576890.5432806, "loss": 2.142350959777832, "grad_norm": 0.337890625, "learning_rate": 0.0007, "train/total_time_seconds": 610.7138260900974, "train/time_per_step_avg": 0.26379979219287636, "train/epoch_time_elapsed": 734.5135913006961, "train/estimated_remaining_minutes": 0.8850925015798513}
116
+ {"step": 2320, "epoch": 0.15638692281766095, "timestamp": 1786576897.167037, "loss": 2.1518985748291017, "grad_norm": 0.333984375, "learning_rate": 0.0007, "train/total_time_seconds": 616.0249392129481, "train/time_per_step_avg": 0.26380566362291574, "train/epoch_time_elapsed": 741.1373476348817, "train/estimated_remaining_minutes": 0.7965839731201914}
117
+ {"step": 2340, "epoch": 0.1577350859453994, "timestamp": 1786576903.46742, "loss": 2.142623710632324, "grad_norm": 0.337890625, "learning_rate": 0.0007, "train/total_time_seconds": 621.279378991574, "train/time_per_step_avg": 0.26364951584488155, "train/epoch_time_elapsed": 747.4377302899957, "train/estimated_remaining_minutes": 0.7080106883094861}
118
+ {"step": 2360, "epoch": 0.15908324907313784, "timestamp": 1786576909.7320642, "loss": 2.1457294464111327, "grad_norm": 0.333984375, "learning_rate": 0.0007, "train/total_time_seconds": 626.5288925841451, "train/time_per_step_avg": 0.26313615657389167, "train/epoch_time_elapsed": 753.7023743912578, "train/estimated_remaining_minutes": 0.6194494700690699}
119
+ {"step": 2380, "epoch": 0.1604314122008763, "timestamp": 1786576915.9947436, "loss": 2.1467227935791016, "grad_norm": 0.33203125, "learning_rate": 0.0007, "train/total_time_seconds": 631.7887535169721, "train/time_per_step_avg": 0.26310893446207045, "train/epoch_time_elapsed": 759.9650537855923, "train/estimated_remaining_minutes": 0.5309149189218253}
120
+ {"step": 2400, "epoch": 0.16177957532861476, "timestamp": 1786576922.265703, "loss": 2.1286346435546877, "grad_norm": 0.369140625, "learning_rate": 0.0007, "train/total_time_seconds": 637.0409996919334, "train/time_per_step_avg": 0.2632717360183597, "train/epoch_time_elapsed": 766.2360133305192, "train/estimated_remaining_minutes": 0.44238958311939824}
121
+ {"step": 2420, "epoch": 0.16312773845635323, "timestamp": 1786576929.1628985, "loss": 2.1322181701660154, "grad_norm": 0.34375, "learning_rate": 0.0007, "train/total_time_seconds": 642.6275409348309, "train/time_per_step_avg": 0.2660260172188282, "train/epoch_time_elapsed": 773.1332089044154, "train/estimated_remaining_minutes": 0.3540647608456369}
122
+ {"step": 2440, "epoch": 0.16447590158409167, "timestamp": 1786576935.4599833, "loss": 2.119257354736328, "grad_norm": 0.330078125, "learning_rate": 0.0007, "train/total_time_seconds": 647.9019580893219, "train/time_per_step_avg": 0.266225790977478, "train/epoch_time_elapsed": 779.43029339239, "train/estimated_remaining_minutes": 0.26553358938086963}
123
+ {"step": 2460, "epoch": 0.16582406471183014, "timestamp": 1786576941.745857, "loss": 2.1314579010009767, "grad_norm": 0.341796875, "learning_rate": 0.0007, "train/total_time_seconds": 653.15468756482, "train/time_per_step_avg": 0.2662579498067498, "train/epoch_time_elapsed": 785.7161669656634, "train/estimated_remaining_minutes": 0.17700669039697023}
124
+ {"step": 2480, "epoch": 0.1671722278395686, "timestamp": 1786576947.9930341, "loss": 2.1324623107910154, "grad_norm": 0.34765625, "learning_rate": 0.0007, "train/total_time_seconds": 658.3901137858629, "train/time_per_step_avg": 0.26601360268890856, "train/epoch_time_elapsed": 791.9633442051709, "train/estimated_remaining_minutes": 0.08849329486369124}
125
+ {"step": 2498, "epoch": 0.1683855746545332, "timestamp": 1786576970.4860284, "eval_loss": 2.125929832458496, "eval_runtime": 16.8275, "eval_samples_per_second": 566.157, "eval_steps_per_second": 0.594, "train/total_time_seconds": 663.1301065795124, "train/time_per_step_avg": 0.2661243563145399, "train/epoch_time_elapsed": 814.4563361778855, "train/estimated_remaining_minutes": 0.008848813805437847}
126
+ {"step": 2500, "epoch": 0.16852039096730703, "timestamp": 1786576971.1115985, "loss": 2.1140785217285156, "grad_norm": 0.333984375, "learning_rate": 0.0007, "train/total_time_seconds": 663.6534794270992, "train/time_per_step_avg": 0.26612479735165834, "train/epoch_time_elapsed": 815.0819086916745, "train/estimated_remaining_minutes": 0.0}
127
+ {"step": 2500, "epoch": 0.16852039096730703, "timestamp": 1786576987.748201, "eval_loss": 2.1262547969818115, "eval_runtime": 16.635, "eval_samples_per_second": 572.71, "eval_steps_per_second": 0.601, "train/total_time_seconds": 663.6534794270992, "train/time_per_step_avg": 0.26612479735165834, "train/epoch_time_elapsed": 831.7185098268092, "train/estimated_remaining_minutes": 0.0}
128
+ {"step": 2500, "epoch": 0.16852039096730703, "timestamp": 1786576988.0397062, "train_runtime": 832.8333, "train_samples_per_second": 192.115, "train_steps_per_second": 3.002, "total_flos": 8065709506560000.0, "train_loss": 3.0849436630249025, "train/total_time_seconds": 663.6534794270992, "train/time_per_step_avg": 0.26612479735165834, "train/epoch_time_elapsed": 832.0100145116448, "train/estimated_remaining_minutes": 0.0}
129
+ {"step": 2500, "epoch": 0.16852039096730703, "timestamp": 1786577004.835671, "eval_loss": 2.1262547969818115, "eval_runtime": 16.7922, "eval_samples_per_second": 567.347, "eval_steps_per_second": 0.596, "train/total_time_seconds": 663.6534794270992, "train/time_per_step_avg": 0.26612479735165834, "train/epoch_time_elapsed": 848.8059792406857, "train/estimated_remaining_minutes": 0.0}