w-ahmad commited on
Commit
d5a4e7d
·
verified ·
1 Parent(s): 25da6e9

Auto upload 2026-08-06T22:11:34.686163 (part 2)

Browse files
Files changed (27) hide show
  1. out/glu-situglu-46L_run/README.md +65 -0
  2. out/glu-situglu-46L_run/checkpoint-1500/config.json +35 -0
  3. out/glu-situglu-46L_run/checkpoint-1500/model.safetensors +3 -0
  4. out/glu-situglu-46L_run/checkpoint-1500/optimizer.pt +3 -0
  5. out/glu-situglu-46L_run/checkpoint-1500/rng_state.pth +3 -0
  6. out/glu-situglu-46L_run/checkpoint-1500/scheduler.pt +3 -0
  7. out/glu-situglu-46L_run/checkpoint-1500/tokenizer.json +0 -0
  8. out/glu-situglu-46L_run/checkpoint-1500/tokenizer_config.json +13 -0
  9. out/glu-situglu-46L_run/checkpoint-1500/trainer_state.json +615 -0
  10. out/glu-situglu-46L_run/checkpoint-1500/training_args.bin +3 -0
  11. out/glu-situglu-46L_run/checkpoint-2000/config.json +35 -0
  12. out/glu-situglu-46L_run/checkpoint-2000/model.safetensors +3 -0
  13. out/glu-situglu-46L_run/checkpoint-2000/optimizer.pt +3 -0
  14. out/glu-situglu-46L_run/checkpoint-2000/rng_state.pth +3 -0
  15. out/glu-situglu-46L_run/checkpoint-2000/scheduler.pt +3 -0
  16. out/glu-situglu-46L_run/checkpoint-2000/tokenizer.json +0 -0
  17. out/glu-situglu-46L_run/checkpoint-2000/tokenizer_config.json +13 -0
  18. out/glu-situglu-46L_run/checkpoint-2000/trainer_state.json +814 -0
  19. out/glu-situglu-46L_run/checkpoint-2000/training_args.bin +3 -0
  20. out/glu-situglu-46L_run/model.safetensors +1 -1
  21. out/glu-situglu-46L_run/tokenizer.json +0 -0
  22. out/glu-situglu-46L_run/tokenizer_config.json +13 -0
  23. out/glu-situglu-46L_run/training_log.jsonl +0 -0
  24. out/sweep_summary.json +4 -4
  25. wandb/debug-internal.log +16 -45
  26. wandb/debug.log +22 -22
  27. wandb/run-20260806_202910-50qq810b/logs/debug-core.log +12 -0
out/glu-situglu-46L_run/README.md ADDED
@@ -0,0 +1,65 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ tags:
4
+ - generated_from_trainer
5
+ model-index:
6
+ - name: ACTio-glu-situglu-46L
7
+ results: []
8
+ ---
9
+
10
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
11
+ should probably proofread and complete it, then remove this comment. -->
12
+
13
+ # ACTio-glu-situglu-46L
14
+
15
+ This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset.
16
+ It achieves the following results on the evaluation set:
17
+ - Loss: 0.2534
18
+
19
+ ## Model description
20
+
21
+ More information needed
22
+
23
+ ## Intended uses & limitations
24
+
25
+ More information needed
26
+
27
+ ## Training and evaluation data
28
+
29
+ More information needed
30
+
31
+ ## Training procedure
32
+
33
+ ### Training hyperparameters
34
+
35
+ The following hyperparameters were used during training:
36
+ - learning_rate: 0.0003
37
+ - train_batch_size: 64
38
+ - eval_batch_size: 64
39
+ - seed: 42
40
+ - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
41
+ - lr_scheduler_type: constant
42
+ - training_steps: 2000
43
+
44
+ ### Training results
45
+
46
+ | Training Loss | Epoch | Step | Validation Loss |
47
+ |:-------------:|:------:|:----:|:---------------:|
48
+ | 4.3064 | 0.0135 | 200 | 4.0854 |
49
+ | 1.6443 | 0.0270 | 400 | 1.5560 |
50
+ | 0.6373 | 0.0404 | 600 | 0.6189 |
51
+ | 0.4392 | 0.0539 | 800 | 0.4295 |
52
+ | 0.3520 | 0.0674 | 1000 | 0.3497 |
53
+ | 0.3347 | 0.0809 | 1200 | 0.3229 |
54
+ | 0.2879 | 0.0944 | 1400 | 0.2899 |
55
+ | 0.3054 | 0.1079 | 1600 | 0.3011 |
56
+ | 0.2637 | 0.1213 | 1800 | 0.2626 |
57
+ | 0.2518 | 0.1348 | 2000 | 0.2534 |
58
+
59
+
60
+ ### Framework versions
61
+
62
+ - Transformers 5.15.0.dev0
63
+ - Pytorch 2.6.0+cu124
64
+ - Datasets 5.0.1
65
+ - Tokenizers 0.22.2
out/glu-situglu-46L_run/checkpoint-1500/config.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "situglu",
3
+ "architectures": [
4
+ "TinyLlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 2,
11
+ "head_dim": 32,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 128,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 256,
16
+ "max_position_embeddings": 512,
17
+ "mlp_bias": false,
18
+ "mlp_type": "glu",
19
+ "model_type": "tiny_llama",
20
+ "num_attention_heads": 4,
21
+ "num_hidden_layers": 46,
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": "5.15.0.dev0",
33
+ "use_cache": false,
34
+ "vocab_size": 4096
35
+ }
out/glu-situglu-46L_run/checkpoint-1500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ed02a3a2a8b40dfd4462cd036af8fa1e3fcb2f9473c00f1d4d403eddfa3438f1
3
+ size 16191392
out/glu-situglu-46L_run/checkpoint-1500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8458feb6edad58929c6865b8117a0521ca1517ced1c2560ab868c888ed35b51b
3
+ size 32639090
out/glu-situglu-46L_run/checkpoint-1500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e2768285b45b2a0c05f6f50bbb8c0287fca6f62a8cde6d1b1f02151ac72ee8dc
3
+ size 14244
out/glu-situglu-46L_run/checkpoint-1500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b26a759cc207e28dea06eae7ad531c2dd78a09978f6972edbfc1566adccd7142
3
+ size 1064
out/glu-situglu-46L_run/checkpoint-1500/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
out/glu-situglu-46L_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
+ }
out/glu-situglu-46L_run/checkpoint-1500/trainer_state.json ADDED
@@ -0,0 +1,615 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.10111223458038422,
6
+ "eval_steps": 200,
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": 1.2578125,
15
+ "learning_rate": 0.0003,
16
+ "loss": 7.777313995361328,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.1640625,
22
+ "learning_rate": 0.0003,
23
+ "loss": 7.018227386474609,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 0.9296875,
29
+ "learning_rate": 0.0003,
30
+ "loss": 6.487169647216797,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 0.68359375,
36
+ "learning_rate": 0.0003,
37
+ "loss": 6.167935562133789,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 0.9296875,
43
+ "learning_rate": 0.0003,
44
+ "loss": 5.984228515625,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 0.71484375,
50
+ "learning_rate": 0.0003,
51
+ "loss": 5.803854370117188,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 0.71484375,
57
+ "learning_rate": 0.0003,
58
+ "loss": 5.414020919799805,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 1.8046875,
64
+ "learning_rate": 0.0003,
65
+ "loss": 5.042311477661133,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 1.15625,
71
+ "learning_rate": 0.0003,
72
+ "loss": 4.7203624725341795,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 1.0078125,
78
+ "learning_rate": 0.0003,
79
+ "loss": 4.306383895874023,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.013481631277384564,
84
+ "eval_loss": 4.0854411125183105,
85
+ "eval_runtime": 12.9528,
86
+ "eval_samples_per_second": 735.516,
87
+ "eval_steps_per_second": 11.503,
88
+ "step": 200
89
+ },
90
+ {
91
+ "epoch": 0.01482979440512302,
92
+ "grad_norm": 1.2578125,
93
+ "learning_rate": 0.0003,
94
+ "loss": 3.9115909576416015,
95
+ "step": 220
96
+ },
97
+ {
98
+ "epoch": 0.016177957532861477,
99
+ "grad_norm": 1.3359375,
100
+ "learning_rate": 0.0003,
101
+ "loss": 3.5526866912841797,
102
+ "step": 240
103
+ },
104
+ {
105
+ "epoch": 0.01752612066059993,
106
+ "grad_norm": 1.3828125,
107
+ "learning_rate": 0.0003,
108
+ "loss": 3.255462646484375,
109
+ "step": 260
110
+ },
111
+ {
112
+ "epoch": 0.01887428378833839,
113
+ "grad_norm": 1.640625,
114
+ "learning_rate": 0.0003,
115
+ "loss": 2.9794519424438475,
116
+ "step": 280
117
+ },
118
+ {
119
+ "epoch": 0.020222446916076844,
120
+ "grad_norm": 1.6328125,
121
+ "learning_rate": 0.0003,
122
+ "loss": 2.7179059982299805,
123
+ "step": 300
124
+ },
125
+ {
126
+ "epoch": 0.021570610043815303,
127
+ "grad_norm": 1.984375,
128
+ "learning_rate": 0.0003,
129
+ "loss": 2.466050720214844,
130
+ "step": 320
131
+ },
132
+ {
133
+ "epoch": 0.022918773171553757,
134
+ "grad_norm": 1.640625,
135
+ "learning_rate": 0.0003,
136
+ "loss": 2.239631462097168,
137
+ "step": 340
138
+ },
139
+ {
140
+ "epoch": 0.024266936299292215,
141
+ "grad_norm": 1.296875,
142
+ "learning_rate": 0.0003,
143
+ "loss": 2.0346961975097657,
144
+ "step": 360
145
+ },
146
+ {
147
+ "epoch": 0.02561509942703067,
148
+ "grad_norm": 1.25,
149
+ "learning_rate": 0.0003,
150
+ "loss": 1.8381683349609375,
151
+ "step": 380
152
+ },
153
+ {
154
+ "epoch": 0.026963262554769128,
155
+ "grad_norm": 1.34375,
156
+ "learning_rate": 0.0003,
157
+ "loss": 1.6442916870117188,
158
+ "step": 400
159
+ },
160
+ {
161
+ "epoch": 0.026963262554769128,
162
+ "eval_loss": 1.5559693574905396,
163
+ "eval_runtime": 12.8417,
164
+ "eval_samples_per_second": 741.882,
165
+ "eval_steps_per_second": 11.603,
166
+ "step": 400
167
+ },
168
+ {
169
+ "epoch": 0.028311425682507583,
170
+ "grad_norm": 1.734375,
171
+ "learning_rate": 0.0003,
172
+ "loss": 1.468704891204834,
173
+ "step": 420
174
+ },
175
+ {
176
+ "epoch": 0.02965958881024604,
177
+ "grad_norm": 0.98046875,
178
+ "learning_rate": 0.0003,
179
+ "loss": 1.3365660667419434,
180
+ "step": 440
181
+ },
182
+ {
183
+ "epoch": 0.031007751937984496,
184
+ "grad_norm": 1.125,
185
+ "learning_rate": 0.0003,
186
+ "loss": 1.1876847267150878,
187
+ "step": 460
188
+ },
189
+ {
190
+ "epoch": 0.032355915065722954,
191
+ "grad_norm": 1.203125,
192
+ "learning_rate": 0.0003,
193
+ "loss": 1.0664135932922363,
194
+ "step": 480
195
+ },
196
+ {
197
+ "epoch": 0.03370407819346141,
198
+ "grad_norm": 0.88671875,
199
+ "learning_rate": 0.0003,
200
+ "loss": 0.9525643348693847,
201
+ "step": 500
202
+ },
203
+ {
204
+ "epoch": 0.03505224132119986,
205
+ "grad_norm": 0.443359375,
206
+ "learning_rate": 0.0003,
207
+ "loss": 0.8542046546936035,
208
+ "step": 520
209
+ },
210
+ {
211
+ "epoch": 0.03640040444893832,
212
+ "grad_norm": 0.478515625,
213
+ "learning_rate": 0.0003,
214
+ "loss": 0.7644352912902832,
215
+ "step": 540
216
+ },
217
+ {
218
+ "epoch": 0.03774856757667678,
219
+ "grad_norm": 0.458984375,
220
+ "learning_rate": 0.0003,
221
+ "loss": 0.7011470794677734,
222
+ "step": 560
223
+ },
224
+ {
225
+ "epoch": 0.03909673070441524,
226
+ "grad_norm": 1.1015625,
227
+ "learning_rate": 0.0003,
228
+ "loss": 0.6564885139465332,
229
+ "step": 580
230
+ },
231
+ {
232
+ "epoch": 0.04044489383215369,
233
+ "grad_norm": 1.125,
234
+ "learning_rate": 0.0003,
235
+ "loss": 0.6372809410095215,
236
+ "step": 600
237
+ },
238
+ {
239
+ "epoch": 0.04044489383215369,
240
+ "eval_loss": 0.6188518404960632,
241
+ "eval_runtime": 12.883,
242
+ "eval_samples_per_second": 739.5,
243
+ "eval_steps_per_second": 11.566,
244
+ "step": 600
245
+ },
246
+ {
247
+ "epoch": 0.04179305695989215,
248
+ "grad_norm": 0.376953125,
249
+ "learning_rate": 0.0003,
250
+ "loss": 0.5965685844421387,
251
+ "step": 620
252
+ },
253
+ {
254
+ "epoch": 0.043141220087630605,
255
+ "grad_norm": 0.31640625,
256
+ "learning_rate": 0.0003,
257
+ "loss": 0.5641400814056396,
258
+ "step": 640
259
+ },
260
+ {
261
+ "epoch": 0.044489383215369056,
262
+ "grad_norm": 0.294921875,
263
+ "learning_rate": 0.0003,
264
+ "loss": 0.5308821678161622,
265
+ "step": 660
266
+ },
267
+ {
268
+ "epoch": 0.045837546343107514,
269
+ "grad_norm": 0.298828125,
270
+ "learning_rate": 0.0003,
271
+ "loss": 0.5098359107971191,
272
+ "step": 680
273
+ },
274
+ {
275
+ "epoch": 0.04718570947084597,
276
+ "grad_norm": 0.31640625,
277
+ "learning_rate": 0.0003,
278
+ "loss": 0.48752245903015134,
279
+ "step": 700
280
+ },
281
+ {
282
+ "epoch": 0.04853387259858443,
283
+ "grad_norm": 0.32421875,
284
+ "learning_rate": 0.0003,
285
+ "loss": 0.47258787155151366,
286
+ "step": 720
287
+ },
288
+ {
289
+ "epoch": 0.04988203572632288,
290
+ "grad_norm": 0.32421875,
291
+ "learning_rate": 0.0003,
292
+ "loss": 0.45342297554016114,
293
+ "step": 740
294
+ },
295
+ {
296
+ "epoch": 0.05123019885406134,
297
+ "grad_norm": 0.376953125,
298
+ "learning_rate": 0.0003,
299
+ "loss": 0.44418673515319823,
300
+ "step": 760
301
+ },
302
+ {
303
+ "epoch": 0.0525783619817998,
304
+ "grad_norm": 1.515625,
305
+ "learning_rate": 0.0003,
306
+ "loss": 0.45928277969360354,
307
+ "step": 780
308
+ },
309
+ {
310
+ "epoch": 0.053926525109538256,
311
+ "grad_norm": 0.703125,
312
+ "learning_rate": 0.0003,
313
+ "loss": 0.4391960620880127,
314
+ "step": 800
315
+ },
316
+ {
317
+ "epoch": 0.053926525109538256,
318
+ "eval_loss": 0.42953070998191833,
319
+ "eval_runtime": 12.9323,
320
+ "eval_samples_per_second": 736.685,
321
+ "eval_steps_per_second": 11.522,
322
+ "step": 800
323
+ },
324
+ {
325
+ "epoch": 0.05527468823727671,
326
+ "grad_norm": 0.3125,
327
+ "learning_rate": 0.0003,
328
+ "loss": 0.4206435203552246,
329
+ "step": 820
330
+ },
331
+ {
332
+ "epoch": 0.056622851365015166,
333
+ "grad_norm": 0.27734375,
334
+ "learning_rate": 0.0003,
335
+ "loss": 0.40812020301818847,
336
+ "step": 840
337
+ },
338
+ {
339
+ "epoch": 0.057971014492753624,
340
+ "grad_norm": 0.265625,
341
+ "learning_rate": 0.0003,
342
+ "loss": 0.39541823863983155,
343
+ "step": 860
344
+ },
345
+ {
346
+ "epoch": 0.05931917762049208,
347
+ "grad_norm": 0.271484375,
348
+ "learning_rate": 0.0003,
349
+ "loss": 0.384733247756958,
350
+ "step": 880
351
+ },
352
+ {
353
+ "epoch": 0.06066734074823053,
354
+ "grad_norm": 0.26171875,
355
+ "learning_rate": 0.0003,
356
+ "loss": 0.38122236728668213,
357
+ "step": 900
358
+ },
359
+ {
360
+ "epoch": 0.06201550387596899,
361
+ "grad_norm": 0.263671875,
362
+ "learning_rate": 0.0003,
363
+ "loss": 0.3741811990737915,
364
+ "step": 920
365
+ },
366
+ {
367
+ "epoch": 0.06336366700370745,
368
+ "grad_norm": 0.263671875,
369
+ "learning_rate": 0.0003,
370
+ "loss": 0.3669325828552246,
371
+ "step": 940
372
+ },
373
+ {
374
+ "epoch": 0.06471183013144591,
375
+ "grad_norm": 0.255859375,
376
+ "learning_rate": 0.0003,
377
+ "loss": 0.35938591957092286,
378
+ "step": 960
379
+ },
380
+ {
381
+ "epoch": 0.06605999325918437,
382
+ "grad_norm": 0.27734375,
383
+ "learning_rate": 0.0003,
384
+ "loss": 0.35484201908111573,
385
+ "step": 980
386
+ },
387
+ {
388
+ "epoch": 0.06740815638692282,
389
+ "grad_norm": 0.28125,
390
+ "learning_rate": 0.0003,
391
+ "loss": 0.3519752025604248,
392
+ "step": 1000
393
+ },
394
+ {
395
+ "epoch": 0.06740815638692282,
396
+ "eval_loss": 0.3496977388858795,
397
+ "eval_runtime": 12.8738,
398
+ "eval_samples_per_second": 740.03,
399
+ "eval_steps_per_second": 11.574,
400
+ "step": 1000
401
+ },
402
+ {
403
+ "epoch": 0.06875631951466127,
404
+ "grad_norm": 0.26953125,
405
+ "learning_rate": 0.0003,
406
+ "loss": 0.34456572532653806,
407
+ "step": 1020
408
+ },
409
+ {
410
+ "epoch": 0.07010448264239973,
411
+ "grad_norm": 0.263671875,
412
+ "learning_rate": 0.0003,
413
+ "loss": 0.33809149265289307,
414
+ "step": 1040
415
+ },
416
+ {
417
+ "epoch": 0.07145264577013818,
418
+ "grad_norm": 0.265625,
419
+ "learning_rate": 0.0003,
420
+ "loss": 0.3359847068786621,
421
+ "step": 1060
422
+ },
423
+ {
424
+ "epoch": 0.07280080889787664,
425
+ "grad_norm": 0.287109375,
426
+ "learning_rate": 0.0003,
427
+ "loss": 0.3295122623443604,
428
+ "step": 1080
429
+ },
430
+ {
431
+ "epoch": 0.0741489720256151,
432
+ "grad_norm": 0.2890625,
433
+ "learning_rate": 0.0003,
434
+ "loss": 0.32773594856262206,
435
+ "step": 1100
436
+ },
437
+ {
438
+ "epoch": 0.07549713515335356,
439
+ "grad_norm": 0.28515625,
440
+ "learning_rate": 0.0003,
441
+ "loss": 0.32281956672668455,
442
+ "step": 1120
443
+ },
444
+ {
445
+ "epoch": 0.07684529828109202,
446
+ "grad_norm": 0.267578125,
447
+ "learning_rate": 0.0003,
448
+ "loss": 0.31684436798095705,
449
+ "step": 1140
450
+ },
451
+ {
452
+ "epoch": 0.07819346140883048,
453
+ "grad_norm": 0.30859375,
454
+ "learning_rate": 0.0003,
455
+ "loss": 0.3159658670425415,
456
+ "step": 1160
457
+ },
458
+ {
459
+ "epoch": 0.07954162453656892,
460
+ "grad_norm": 1.7265625,
461
+ "learning_rate": 0.0003,
462
+ "loss": 0.3564677476882935,
463
+ "step": 1180
464
+ },
465
+ {
466
+ "epoch": 0.08088978766430738,
467
+ "grad_norm": 0.5078125,
468
+ "learning_rate": 0.0003,
469
+ "loss": 0.33468797206878664,
470
+ "step": 1200
471
+ },
472
+ {
473
+ "epoch": 0.08088978766430738,
474
+ "eval_loss": 0.32294711470603943,
475
+ "eval_runtime": 12.8444,
476
+ "eval_samples_per_second": 741.724,
477
+ "eval_steps_per_second": 11.6,
478
+ "step": 1200
479
+ },
480
+ {
481
+ "epoch": 0.08223795079204584,
482
+ "grad_norm": 0.283203125,
483
+ "learning_rate": 0.0003,
484
+ "loss": 0.3153939723968506,
485
+ "step": 1220
486
+ },
487
+ {
488
+ "epoch": 0.0835861139197843,
489
+ "grad_norm": 0.25,
490
+ "learning_rate": 0.0003,
491
+ "loss": 0.31013171672821044,
492
+ "step": 1240
493
+ },
494
+ {
495
+ "epoch": 0.08493427704752275,
496
+ "grad_norm": 0.25,
497
+ "learning_rate": 0.0003,
498
+ "loss": 0.30731561183929446,
499
+ "step": 1260
500
+ },
501
+ {
502
+ "epoch": 0.08628244017526121,
503
+ "grad_norm": 0.251953125,
504
+ "learning_rate": 0.0003,
505
+ "loss": 0.30195567607879636,
506
+ "step": 1280
507
+ },
508
+ {
509
+ "epoch": 0.08763060330299967,
510
+ "grad_norm": 0.26171875,
511
+ "learning_rate": 0.0003,
512
+ "loss": 0.30084705352783203,
513
+ "step": 1300
514
+ },
515
+ {
516
+ "epoch": 0.08897876643073811,
517
+ "grad_norm": 0.2421875,
518
+ "learning_rate": 0.0003,
519
+ "loss": 0.297377610206604,
520
+ "step": 1320
521
+ },
522
+ {
523
+ "epoch": 0.09032692955847657,
524
+ "grad_norm": 0.24609375,
525
+ "learning_rate": 0.0003,
526
+ "loss": 0.29534096717834474,
527
+ "step": 1340
528
+ },
529
+ {
530
+ "epoch": 0.09167509268621503,
531
+ "grad_norm": 0.2451171875,
532
+ "learning_rate": 0.0003,
533
+ "loss": 0.29182398319244385,
534
+ "step": 1360
535
+ },
536
+ {
537
+ "epoch": 0.09302325581395349,
538
+ "grad_norm": 0.248046875,
539
+ "learning_rate": 0.0003,
540
+ "loss": 0.2910054922103882,
541
+ "step": 1380
542
+ },
543
+ {
544
+ "epoch": 0.09437141894169195,
545
+ "grad_norm": 0.26171875,
546
+ "learning_rate": 0.0003,
547
+ "loss": 0.2878779172897339,
548
+ "step": 1400
549
+ },
550
+ {
551
+ "epoch": 0.09437141894169195,
552
+ "eval_loss": 0.28991881012916565,
553
+ "eval_runtime": 12.8914,
554
+ "eval_samples_per_second": 739.02,
555
+ "eval_steps_per_second": 11.558,
556
+ "step": 1400
557
+ },
558
+ {
559
+ "epoch": 0.0957195820694304,
560
+ "grad_norm": 0.251953125,
561
+ "learning_rate": 0.0003,
562
+ "loss": 0.2875370502471924,
563
+ "step": 1420
564
+ },
565
+ {
566
+ "epoch": 0.09706774519716886,
567
+ "grad_norm": 0.24609375,
568
+ "learning_rate": 0.0003,
569
+ "loss": 0.28477251529693604,
570
+ "step": 1440
571
+ },
572
+ {
573
+ "epoch": 0.09841590832490732,
574
+ "grad_norm": 0.251953125,
575
+ "learning_rate": 0.0003,
576
+ "loss": 0.28224821090698243,
577
+ "step": 1460
578
+ },
579
+ {
580
+ "epoch": 0.09976407145264576,
581
+ "grad_norm": 0.24609375,
582
+ "learning_rate": 0.0003,
583
+ "loss": 0.2802304744720459,
584
+ "step": 1480
585
+ },
586
+ {
587
+ "epoch": 0.10111223458038422,
588
+ "grad_norm": 0.24609375,
589
+ "learning_rate": 0.0003,
590
+ "loss": 0.27904813289642333,
591
+ "step": 1500
592
+ }
593
+ ],
594
+ "logging_steps": 20,
595
+ "max_steps": 2000,
596
+ "num_input_tokens_seen": 0,
597
+ "num_train_epochs": 1,
598
+ "save_steps": 500,
599
+ "stateful_callbacks": {
600
+ "TrainerControl": {
601
+ "args": {
602
+ "should_epoch_stop": false,
603
+ "should_evaluate": false,
604
+ "should_log": false,
605
+ "should_save": true,
606
+ "should_training_stop": false
607
+ },
608
+ "attributes": {}
609
+ }
610
+ },
611
+ "total_flos": 2226156208128000.0,
612
+ "train_batch_size": 64,
613
+ "trial_name": null,
614
+ "trial_params": null
615
+ }
out/glu-situglu-46L_run/checkpoint-1500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:34907e35fb7d1b911fb670e6ed585e70819d6d19255391284922f26326637bab
3
+ size 4920
out/glu-situglu-46L_run/checkpoint-2000/config.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "situglu",
3
+ "architectures": [
4
+ "TinyLlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 2,
11
+ "head_dim": 32,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 128,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 256,
16
+ "max_position_embeddings": 512,
17
+ "mlp_bias": false,
18
+ "mlp_type": "glu",
19
+ "model_type": "tiny_llama",
20
+ "num_attention_heads": 4,
21
+ "num_hidden_layers": 46,
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": "5.15.0.dev0",
33
+ "use_cache": false,
34
+ "vocab_size": 4096
35
+ }
out/glu-situglu-46L_run/checkpoint-2000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:56a2dfa9ddb1646bec4c23a342aefa2df694d678c02484bddd669a0c3809b1f1
3
+ size 16191392
out/glu-situglu-46L_run/checkpoint-2000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc806169687adf5f417b0dc78e7f34e1731287ae6a6bb20e112c772fcaedd0f0
3
+ size 32639090
out/glu-situglu-46L_run/checkpoint-2000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b1a97db8e41139aa1239ba7fb79ddeb0af5998c6305a440c1fe182e6ad02f2f5
3
+ size 14244
out/glu-situglu-46L_run/checkpoint-2000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e91519ad666af6acc4beafd44aa617aefb90f632388a0f75cbf88f957400c190
3
+ size 1064
out/glu-situglu-46L_run/checkpoint-2000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
out/glu-situglu-46L_run/checkpoint-2000/tokenizer_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|endoftext|>",
5
+ "eos_token": "<|endoftext|>",
6
+ "errors": "replace",
7
+ "is_local": false,
8
+ "local_files_only": false,
9
+ "model_max_length": 1024,
10
+ "pad_token": "<|endoftext|>",
11
+ "tokenizer_class": "GPT2Tokenizer",
12
+ "unk_token": "<|endoftext|>"
13
+ }
out/glu-situglu-46L_run/checkpoint-2000/trainer_state.json ADDED
@@ -0,0 +1,814 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.13481631277384565,
6
+ "eval_steps": 200,
7
+ "global_step": 2000,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0013481631277384564,
14
+ "grad_norm": 1.2578125,
15
+ "learning_rate": 0.0003,
16
+ "loss": 7.777313995361328,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.1640625,
22
+ "learning_rate": 0.0003,
23
+ "loss": 7.018227386474609,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 0.9296875,
29
+ "learning_rate": 0.0003,
30
+ "loss": 6.487169647216797,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 0.68359375,
36
+ "learning_rate": 0.0003,
37
+ "loss": 6.167935562133789,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 0.9296875,
43
+ "learning_rate": 0.0003,
44
+ "loss": 5.984228515625,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 0.71484375,
50
+ "learning_rate": 0.0003,
51
+ "loss": 5.803854370117188,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 0.71484375,
57
+ "learning_rate": 0.0003,
58
+ "loss": 5.414020919799805,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 1.8046875,
64
+ "learning_rate": 0.0003,
65
+ "loss": 5.042311477661133,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 1.15625,
71
+ "learning_rate": 0.0003,
72
+ "loss": 4.7203624725341795,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 1.0078125,
78
+ "learning_rate": 0.0003,
79
+ "loss": 4.306383895874023,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.013481631277384564,
84
+ "eval_loss": 4.0854411125183105,
85
+ "eval_runtime": 12.9528,
86
+ "eval_samples_per_second": 735.516,
87
+ "eval_steps_per_second": 11.503,
88
+ "step": 200
89
+ },
90
+ {
91
+ "epoch": 0.01482979440512302,
92
+ "grad_norm": 1.2578125,
93
+ "learning_rate": 0.0003,
94
+ "loss": 3.9115909576416015,
95
+ "step": 220
96
+ },
97
+ {
98
+ "epoch": 0.016177957532861477,
99
+ "grad_norm": 1.3359375,
100
+ "learning_rate": 0.0003,
101
+ "loss": 3.5526866912841797,
102
+ "step": 240
103
+ },
104
+ {
105
+ "epoch": 0.01752612066059993,
106
+ "grad_norm": 1.3828125,
107
+ "learning_rate": 0.0003,
108
+ "loss": 3.255462646484375,
109
+ "step": 260
110
+ },
111
+ {
112
+ "epoch": 0.01887428378833839,
113
+ "grad_norm": 1.640625,
114
+ "learning_rate": 0.0003,
115
+ "loss": 2.9794519424438475,
116
+ "step": 280
117
+ },
118
+ {
119
+ "epoch": 0.020222446916076844,
120
+ "grad_norm": 1.6328125,
121
+ "learning_rate": 0.0003,
122
+ "loss": 2.7179059982299805,
123
+ "step": 300
124
+ },
125
+ {
126
+ "epoch": 0.021570610043815303,
127
+ "grad_norm": 1.984375,
128
+ "learning_rate": 0.0003,
129
+ "loss": 2.466050720214844,
130
+ "step": 320
131
+ },
132
+ {
133
+ "epoch": 0.022918773171553757,
134
+ "grad_norm": 1.640625,
135
+ "learning_rate": 0.0003,
136
+ "loss": 2.239631462097168,
137
+ "step": 340
138
+ },
139
+ {
140
+ "epoch": 0.024266936299292215,
141
+ "grad_norm": 1.296875,
142
+ "learning_rate": 0.0003,
143
+ "loss": 2.0346961975097657,
144
+ "step": 360
145
+ },
146
+ {
147
+ "epoch": 0.02561509942703067,
148
+ "grad_norm": 1.25,
149
+ "learning_rate": 0.0003,
150
+ "loss": 1.8381683349609375,
151
+ "step": 380
152
+ },
153
+ {
154
+ "epoch": 0.026963262554769128,
155
+ "grad_norm": 1.34375,
156
+ "learning_rate": 0.0003,
157
+ "loss": 1.6442916870117188,
158
+ "step": 400
159
+ },
160
+ {
161
+ "epoch": 0.026963262554769128,
162
+ "eval_loss": 1.5559693574905396,
163
+ "eval_runtime": 12.8417,
164
+ "eval_samples_per_second": 741.882,
165
+ "eval_steps_per_second": 11.603,
166
+ "step": 400
167
+ },
168
+ {
169
+ "epoch": 0.028311425682507583,
170
+ "grad_norm": 1.734375,
171
+ "learning_rate": 0.0003,
172
+ "loss": 1.468704891204834,
173
+ "step": 420
174
+ },
175
+ {
176
+ "epoch": 0.02965958881024604,
177
+ "grad_norm": 0.98046875,
178
+ "learning_rate": 0.0003,
179
+ "loss": 1.3365660667419434,
180
+ "step": 440
181
+ },
182
+ {
183
+ "epoch": 0.031007751937984496,
184
+ "grad_norm": 1.125,
185
+ "learning_rate": 0.0003,
186
+ "loss": 1.1876847267150878,
187
+ "step": 460
188
+ },
189
+ {
190
+ "epoch": 0.032355915065722954,
191
+ "grad_norm": 1.203125,
192
+ "learning_rate": 0.0003,
193
+ "loss": 1.0664135932922363,
194
+ "step": 480
195
+ },
196
+ {
197
+ "epoch": 0.03370407819346141,
198
+ "grad_norm": 0.88671875,
199
+ "learning_rate": 0.0003,
200
+ "loss": 0.9525643348693847,
201
+ "step": 500
202
+ },
203
+ {
204
+ "epoch": 0.03505224132119986,
205
+ "grad_norm": 0.443359375,
206
+ "learning_rate": 0.0003,
207
+ "loss": 0.8542046546936035,
208
+ "step": 520
209
+ },
210
+ {
211
+ "epoch": 0.03640040444893832,
212
+ "grad_norm": 0.478515625,
213
+ "learning_rate": 0.0003,
214
+ "loss": 0.7644352912902832,
215
+ "step": 540
216
+ },
217
+ {
218
+ "epoch": 0.03774856757667678,
219
+ "grad_norm": 0.458984375,
220
+ "learning_rate": 0.0003,
221
+ "loss": 0.7011470794677734,
222
+ "step": 560
223
+ },
224
+ {
225
+ "epoch": 0.03909673070441524,
226
+ "grad_norm": 1.1015625,
227
+ "learning_rate": 0.0003,
228
+ "loss": 0.6564885139465332,
229
+ "step": 580
230
+ },
231
+ {
232
+ "epoch": 0.04044489383215369,
233
+ "grad_norm": 1.125,
234
+ "learning_rate": 0.0003,
235
+ "loss": 0.6372809410095215,
236
+ "step": 600
237
+ },
238
+ {
239
+ "epoch": 0.04044489383215369,
240
+ "eval_loss": 0.6188518404960632,
241
+ "eval_runtime": 12.883,
242
+ "eval_samples_per_second": 739.5,
243
+ "eval_steps_per_second": 11.566,
244
+ "step": 600
245
+ },
246
+ {
247
+ "epoch": 0.04179305695989215,
248
+ "grad_norm": 0.376953125,
249
+ "learning_rate": 0.0003,
250
+ "loss": 0.5965685844421387,
251
+ "step": 620
252
+ },
253
+ {
254
+ "epoch": 0.043141220087630605,
255
+ "grad_norm": 0.31640625,
256
+ "learning_rate": 0.0003,
257
+ "loss": 0.5641400814056396,
258
+ "step": 640
259
+ },
260
+ {
261
+ "epoch": 0.044489383215369056,
262
+ "grad_norm": 0.294921875,
263
+ "learning_rate": 0.0003,
264
+ "loss": 0.5308821678161622,
265
+ "step": 660
266
+ },
267
+ {
268
+ "epoch": 0.045837546343107514,
269
+ "grad_norm": 0.298828125,
270
+ "learning_rate": 0.0003,
271
+ "loss": 0.5098359107971191,
272
+ "step": 680
273
+ },
274
+ {
275
+ "epoch": 0.04718570947084597,
276
+ "grad_norm": 0.31640625,
277
+ "learning_rate": 0.0003,
278
+ "loss": 0.48752245903015134,
279
+ "step": 700
280
+ },
281
+ {
282
+ "epoch": 0.04853387259858443,
283
+ "grad_norm": 0.32421875,
284
+ "learning_rate": 0.0003,
285
+ "loss": 0.47258787155151366,
286
+ "step": 720
287
+ },
288
+ {
289
+ "epoch": 0.04988203572632288,
290
+ "grad_norm": 0.32421875,
291
+ "learning_rate": 0.0003,
292
+ "loss": 0.45342297554016114,
293
+ "step": 740
294
+ },
295
+ {
296
+ "epoch": 0.05123019885406134,
297
+ "grad_norm": 0.376953125,
298
+ "learning_rate": 0.0003,
299
+ "loss": 0.44418673515319823,
300
+ "step": 760
301
+ },
302
+ {
303
+ "epoch": 0.0525783619817998,
304
+ "grad_norm": 1.515625,
305
+ "learning_rate": 0.0003,
306
+ "loss": 0.45928277969360354,
307
+ "step": 780
308
+ },
309
+ {
310
+ "epoch": 0.053926525109538256,
311
+ "grad_norm": 0.703125,
312
+ "learning_rate": 0.0003,
313
+ "loss": 0.4391960620880127,
314
+ "step": 800
315
+ },
316
+ {
317
+ "epoch": 0.053926525109538256,
318
+ "eval_loss": 0.42953070998191833,
319
+ "eval_runtime": 12.9323,
320
+ "eval_samples_per_second": 736.685,
321
+ "eval_steps_per_second": 11.522,
322
+ "step": 800
323
+ },
324
+ {
325
+ "epoch": 0.05527468823727671,
326
+ "grad_norm": 0.3125,
327
+ "learning_rate": 0.0003,
328
+ "loss": 0.4206435203552246,
329
+ "step": 820
330
+ },
331
+ {
332
+ "epoch": 0.056622851365015166,
333
+ "grad_norm": 0.27734375,
334
+ "learning_rate": 0.0003,
335
+ "loss": 0.40812020301818847,
336
+ "step": 840
337
+ },
338
+ {
339
+ "epoch": 0.057971014492753624,
340
+ "grad_norm": 0.265625,
341
+ "learning_rate": 0.0003,
342
+ "loss": 0.39541823863983155,
343
+ "step": 860
344
+ },
345
+ {
346
+ "epoch": 0.05931917762049208,
347
+ "grad_norm": 0.271484375,
348
+ "learning_rate": 0.0003,
349
+ "loss": 0.384733247756958,
350
+ "step": 880
351
+ },
352
+ {
353
+ "epoch": 0.06066734074823053,
354
+ "grad_norm": 0.26171875,
355
+ "learning_rate": 0.0003,
356
+ "loss": 0.38122236728668213,
357
+ "step": 900
358
+ },
359
+ {
360
+ "epoch": 0.06201550387596899,
361
+ "grad_norm": 0.263671875,
362
+ "learning_rate": 0.0003,
363
+ "loss": 0.3741811990737915,
364
+ "step": 920
365
+ },
366
+ {
367
+ "epoch": 0.06336366700370745,
368
+ "grad_norm": 0.263671875,
369
+ "learning_rate": 0.0003,
370
+ "loss": 0.3669325828552246,
371
+ "step": 940
372
+ },
373
+ {
374
+ "epoch": 0.06471183013144591,
375
+ "grad_norm": 0.255859375,
376
+ "learning_rate": 0.0003,
377
+ "loss": 0.35938591957092286,
378
+ "step": 960
379
+ },
380
+ {
381
+ "epoch": 0.06605999325918437,
382
+ "grad_norm": 0.27734375,
383
+ "learning_rate": 0.0003,
384
+ "loss": 0.35484201908111573,
385
+ "step": 980
386
+ },
387
+ {
388
+ "epoch": 0.06740815638692282,
389
+ "grad_norm": 0.28125,
390
+ "learning_rate": 0.0003,
391
+ "loss": 0.3519752025604248,
392
+ "step": 1000
393
+ },
394
+ {
395
+ "epoch": 0.06740815638692282,
396
+ "eval_loss": 0.3496977388858795,
397
+ "eval_runtime": 12.8738,
398
+ "eval_samples_per_second": 740.03,
399
+ "eval_steps_per_second": 11.574,
400
+ "step": 1000
401
+ },
402
+ {
403
+ "epoch": 0.06875631951466127,
404
+ "grad_norm": 0.26953125,
405
+ "learning_rate": 0.0003,
406
+ "loss": 0.34456572532653806,
407
+ "step": 1020
408
+ },
409
+ {
410
+ "epoch": 0.07010448264239973,
411
+ "grad_norm": 0.263671875,
412
+ "learning_rate": 0.0003,
413
+ "loss": 0.33809149265289307,
414
+ "step": 1040
415
+ },
416
+ {
417
+ "epoch": 0.07145264577013818,
418
+ "grad_norm": 0.265625,
419
+ "learning_rate": 0.0003,
420
+ "loss": 0.3359847068786621,
421
+ "step": 1060
422
+ },
423
+ {
424
+ "epoch": 0.07280080889787664,
425
+ "grad_norm": 0.287109375,
426
+ "learning_rate": 0.0003,
427
+ "loss": 0.3295122623443604,
428
+ "step": 1080
429
+ },
430
+ {
431
+ "epoch": 0.0741489720256151,
432
+ "grad_norm": 0.2890625,
433
+ "learning_rate": 0.0003,
434
+ "loss": 0.32773594856262206,
435
+ "step": 1100
436
+ },
437
+ {
438
+ "epoch": 0.07549713515335356,
439
+ "grad_norm": 0.28515625,
440
+ "learning_rate": 0.0003,
441
+ "loss": 0.32281956672668455,
442
+ "step": 1120
443
+ },
444
+ {
445
+ "epoch": 0.07684529828109202,
446
+ "grad_norm": 0.267578125,
447
+ "learning_rate": 0.0003,
448
+ "loss": 0.31684436798095705,
449
+ "step": 1140
450
+ },
451
+ {
452
+ "epoch": 0.07819346140883048,
453
+ "grad_norm": 0.30859375,
454
+ "learning_rate": 0.0003,
455
+ "loss": 0.3159658670425415,
456
+ "step": 1160
457
+ },
458
+ {
459
+ "epoch": 0.07954162453656892,
460
+ "grad_norm": 1.7265625,
461
+ "learning_rate": 0.0003,
462
+ "loss": 0.3564677476882935,
463
+ "step": 1180
464
+ },
465
+ {
466
+ "epoch": 0.08088978766430738,
467
+ "grad_norm": 0.5078125,
468
+ "learning_rate": 0.0003,
469
+ "loss": 0.33468797206878664,
470
+ "step": 1200
471
+ },
472
+ {
473
+ "epoch": 0.08088978766430738,
474
+ "eval_loss": 0.32294711470603943,
475
+ "eval_runtime": 12.8444,
476
+ "eval_samples_per_second": 741.724,
477
+ "eval_steps_per_second": 11.6,
478
+ "step": 1200
479
+ },
480
+ {
481
+ "epoch": 0.08223795079204584,
482
+ "grad_norm": 0.283203125,
483
+ "learning_rate": 0.0003,
484
+ "loss": 0.3153939723968506,
485
+ "step": 1220
486
+ },
487
+ {
488
+ "epoch": 0.0835861139197843,
489
+ "grad_norm": 0.25,
490
+ "learning_rate": 0.0003,
491
+ "loss": 0.31013171672821044,
492
+ "step": 1240
493
+ },
494
+ {
495
+ "epoch": 0.08493427704752275,
496
+ "grad_norm": 0.25,
497
+ "learning_rate": 0.0003,
498
+ "loss": 0.30731561183929446,
499
+ "step": 1260
500
+ },
501
+ {
502
+ "epoch": 0.08628244017526121,
503
+ "grad_norm": 0.251953125,
504
+ "learning_rate": 0.0003,
505
+ "loss": 0.30195567607879636,
506
+ "step": 1280
507
+ },
508
+ {
509
+ "epoch": 0.08763060330299967,
510
+ "grad_norm": 0.26171875,
511
+ "learning_rate": 0.0003,
512
+ "loss": 0.30084705352783203,
513
+ "step": 1300
514
+ },
515
+ {
516
+ "epoch": 0.08897876643073811,
517
+ "grad_norm": 0.2421875,
518
+ "learning_rate": 0.0003,
519
+ "loss": 0.297377610206604,
520
+ "step": 1320
521
+ },
522
+ {
523
+ "epoch": 0.09032692955847657,
524
+ "grad_norm": 0.24609375,
525
+ "learning_rate": 0.0003,
526
+ "loss": 0.29534096717834474,
527
+ "step": 1340
528
+ },
529
+ {
530
+ "epoch": 0.09167509268621503,
531
+ "grad_norm": 0.2451171875,
532
+ "learning_rate": 0.0003,
533
+ "loss": 0.29182398319244385,
534
+ "step": 1360
535
+ },
536
+ {
537
+ "epoch": 0.09302325581395349,
538
+ "grad_norm": 0.248046875,
539
+ "learning_rate": 0.0003,
540
+ "loss": 0.2910054922103882,
541
+ "step": 1380
542
+ },
543
+ {
544
+ "epoch": 0.09437141894169195,
545
+ "grad_norm": 0.26171875,
546
+ "learning_rate": 0.0003,
547
+ "loss": 0.2878779172897339,
548
+ "step": 1400
549
+ },
550
+ {
551
+ "epoch": 0.09437141894169195,
552
+ "eval_loss": 0.28991881012916565,
553
+ "eval_runtime": 12.8914,
554
+ "eval_samples_per_second": 739.02,
555
+ "eval_steps_per_second": 11.558,
556
+ "step": 1400
557
+ },
558
+ {
559
+ "epoch": 0.0957195820694304,
560
+ "grad_norm": 0.251953125,
561
+ "learning_rate": 0.0003,
562
+ "loss": 0.2875370502471924,
563
+ "step": 1420
564
+ },
565
+ {
566
+ "epoch": 0.09706774519716886,
567
+ "grad_norm": 0.24609375,
568
+ "learning_rate": 0.0003,
569
+ "loss": 0.28477251529693604,
570
+ "step": 1440
571
+ },
572
+ {
573
+ "epoch": 0.09841590832490732,
574
+ "grad_norm": 0.251953125,
575
+ "learning_rate": 0.0003,
576
+ "loss": 0.28224821090698243,
577
+ "step": 1460
578
+ },
579
+ {
580
+ "epoch": 0.09976407145264576,
581
+ "grad_norm": 0.24609375,
582
+ "learning_rate": 0.0003,
583
+ "loss": 0.2802304744720459,
584
+ "step": 1480
585
+ },
586
+ {
587
+ "epoch": 0.10111223458038422,
588
+ "grad_norm": 0.24609375,
589
+ "learning_rate": 0.0003,
590
+ "loss": 0.27904813289642333,
591
+ "step": 1500
592
+ },
593
+ {
594
+ "epoch": 0.10246039770812268,
595
+ "grad_norm": 0.248046875,
596
+ "learning_rate": 0.0003,
597
+ "loss": 0.27666282653808594,
598
+ "step": 1520
599
+ },
600
+ {
601
+ "epoch": 0.10380856083586114,
602
+ "grad_norm": 0.259765625,
603
+ "learning_rate": 0.0003,
604
+ "loss": 0.2755317211151123,
605
+ "step": 1540
606
+ },
607
+ {
608
+ "epoch": 0.1051567239635996,
609
+ "grad_norm": 0.25390625,
610
+ "learning_rate": 0.0003,
611
+ "loss": 0.27470548152923585,
612
+ "step": 1560
613
+ },
614
+ {
615
+ "epoch": 0.10650488709133805,
616
+ "grad_norm": 0.259765625,
617
+ "learning_rate": 0.0003,
618
+ "loss": 0.27537152767181394,
619
+ "step": 1580
620
+ },
621
+ {
622
+ "epoch": 0.10785305021907651,
623
+ "grad_norm": 1.5703125,
624
+ "learning_rate": 0.0003,
625
+ "loss": 0.3053688287734985,
626
+ "step": 1600
627
+ },
628
+ {
629
+ "epoch": 0.10785305021907651,
630
+ "eval_loss": 0.3011101484298706,
631
+ "eval_runtime": 12.9063,
632
+ "eval_samples_per_second": 738.168,
633
+ "eval_steps_per_second": 11.545,
634
+ "step": 1600
635
+ },
636
+ {
637
+ "epoch": 0.10920121334681497,
638
+ "grad_norm": 0.45703125,
639
+ "learning_rate": 0.0003,
640
+ "loss": 0.28953573703765867,
641
+ "step": 1620
642
+ },
643
+ {
644
+ "epoch": 0.11054937647455342,
645
+ "grad_norm": 0.265625,
646
+ "learning_rate": 0.0003,
647
+ "loss": 0.2765570878982544,
648
+ "step": 1640
649
+ },
650
+ {
651
+ "epoch": 0.11189753960229187,
652
+ "grad_norm": 0.240234375,
653
+ "learning_rate": 0.0003,
654
+ "loss": 0.2706732749938965,
655
+ "step": 1660
656
+ },
657
+ {
658
+ "epoch": 0.11324570273003033,
659
+ "grad_norm": 0.234375,
660
+ "learning_rate": 0.0003,
661
+ "loss": 0.26884875297546384,
662
+ "step": 1680
663
+ },
664
+ {
665
+ "epoch": 0.11459386585776879,
666
+ "grad_norm": 0.236328125,
667
+ "learning_rate": 0.0003,
668
+ "loss": 0.2686060190200806,
669
+ "step": 1700
670
+ },
671
+ {
672
+ "epoch": 0.11594202898550725,
673
+ "grad_norm": 0.2373046875,
674
+ "learning_rate": 0.0003,
675
+ "loss": 0.2660403251647949,
676
+ "step": 1720
677
+ },
678
+ {
679
+ "epoch": 0.1172901921132457,
680
+ "grad_norm": 0.2373046875,
681
+ "learning_rate": 0.0003,
682
+ "loss": 0.26566119194030763,
683
+ "step": 1740
684
+ },
685
+ {
686
+ "epoch": 0.11863835524098416,
687
+ "grad_norm": 0.236328125,
688
+ "learning_rate": 0.0003,
689
+ "loss": 0.26347191333770753,
690
+ "step": 1760
691
+ },
692
+ {
693
+ "epoch": 0.11998651836872262,
694
+ "grad_norm": 0.232421875,
695
+ "learning_rate": 0.0003,
696
+ "loss": 0.26403193473815917,
697
+ "step": 1780
698
+ },
699
+ {
700
+ "epoch": 0.12133468149646107,
701
+ "grad_norm": 0.2333984375,
702
+ "learning_rate": 0.0003,
703
+ "loss": 0.26373121738433836,
704
+ "step": 1800
705
+ },
706
+ {
707
+ "epoch": 0.12133468149646107,
708
+ "eval_loss": 0.2626025974750519,
709
+ "eval_runtime": 12.8738,
710
+ "eval_samples_per_second": 740.029,
711
+ "eval_steps_per_second": 11.574,
712
+ "step": 1800
713
+ },
714
+ {
715
+ "epoch": 0.12268284462419952,
716
+ "grad_norm": 0.232421875,
717
+ "learning_rate": 0.0003,
718
+ "loss": 0.2631482362747192,
719
+ "step": 1820
720
+ },
721
+ {
722
+ "epoch": 0.12403100775193798,
723
+ "grad_norm": 0.2294921875,
724
+ "learning_rate": 0.0003,
725
+ "loss": 0.25759310722351075,
726
+ "step": 1840
727
+ },
728
+ {
729
+ "epoch": 0.12537917087967643,
730
+ "grad_norm": 0.2353515625,
731
+ "learning_rate": 0.0003,
732
+ "loss": 0.2573524475097656,
733
+ "step": 1860
734
+ },
735
+ {
736
+ "epoch": 0.1267273340074149,
737
+ "grad_norm": 0.236328125,
738
+ "learning_rate": 0.0003,
739
+ "loss": 0.256460165977478,
740
+ "step": 1880
741
+ },
742
+ {
743
+ "epoch": 0.12807549713515334,
744
+ "grad_norm": 0.232421875,
745
+ "learning_rate": 0.0003,
746
+ "loss": 0.2563998937606812,
747
+ "step": 1900
748
+ },
749
+ {
750
+ "epoch": 0.12942366026289182,
751
+ "grad_norm": 0.2353515625,
752
+ "learning_rate": 0.0003,
753
+ "loss": 0.25412280559539796,
754
+ "step": 1920
755
+ },
756
+ {
757
+ "epoch": 0.13077182339063026,
758
+ "grad_norm": 0.2333984375,
759
+ "learning_rate": 0.0003,
760
+ "loss": 0.25445094108581545,
761
+ "step": 1940
762
+ },
763
+ {
764
+ "epoch": 0.13211998651836873,
765
+ "grad_norm": 0.228515625,
766
+ "learning_rate": 0.0003,
767
+ "loss": 0.251305890083313,
768
+ "step": 1960
769
+ },
770
+ {
771
+ "epoch": 0.13346814964610718,
772
+ "grad_norm": 0.23046875,
773
+ "learning_rate": 0.0003,
774
+ "loss": 0.2525143623352051,
775
+ "step": 1980
776
+ },
777
+ {
778
+ "epoch": 0.13481631277384565,
779
+ "grad_norm": 0.2275390625,
780
+ "learning_rate": 0.0003,
781
+ "loss": 0.2518260717391968,
782
+ "step": 2000
783
+ },
784
+ {
785
+ "epoch": 0.13481631277384565,
786
+ "eval_loss": 0.2533577084541321,
787
+ "eval_runtime": 12.8891,
788
+ "eval_samples_per_second": 739.153,
789
+ "eval_steps_per_second": 11.56,
790
+ "step": 2000
791
+ }
792
+ ],
793
+ "logging_steps": 20,
794
+ "max_steps": 2000,
795
+ "num_input_tokens_seen": 0,
796
+ "num_train_epochs": 1,
797
+ "save_steps": 500,
798
+ "stateful_callbacks": {
799
+ "TrainerControl": {
800
+ "args": {
801
+ "should_epoch_stop": false,
802
+ "should_evaluate": false,
803
+ "should_log": false,
804
+ "should_save": true,
805
+ "should_training_stop": true
806
+ },
807
+ "attributes": {}
808
+ }
809
+ },
810
+ "total_flos": 2968208277504000.0,
811
+ "train_batch_size": 64,
812
+ "trial_name": null,
813
+ "trial_params": null
814
+ }
out/glu-situglu-46L_run/checkpoint-2000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:34907e35fb7d1b911fb670e6ed585e70819d6d19255391284922f26326637bab
3
+ size 4920
out/glu-situglu-46L_run/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:658ac2ed14a2ccc021fdc6569557ea6acb491a61da931a8b2276383095f73569
3
  size 16191392
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:56a2dfa9ddb1646bec4c23a342aefa2df694d678c02484bddd669a0c3809b1f1
3
  size 16191392
out/glu-situglu-46L_run/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
out/glu-situglu-46L_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
+ }
out/glu-situglu-46L_run/training_log.jsonl CHANGED
The diff for this file is too large to render. See raw diff
 
out/sweep_summary.json CHANGED
@@ -1,8 +1,8 @@
1
  [
2
  {
3
- "variant": "glu-linear-46L",
4
- "eval_loss": 0.23021633923053741,
5
- "out": "out/glu-linear-46L_run",
6
- "run_name": "LM-glu-linear-46L-8.1M-20260806-215401"
7
  }
8
  ]
 
1
  [
2
  {
3
+ "variant": "glu-situglu-46L",
4
+ "eval_loss": 0.2533577084541321,
5
+ "out": "out/glu-situglu-46L_run",
6
+ "run_name": "LM-glu-situglu-46L-8.1M-20260806-220208"
7
  }
8
  ]
wandb/debug-internal.log CHANGED
@@ -1,45 +1,16 @@
1
- {"time":"2026-08-06T22:02:10.439268987Z","level":"INFO","msg":"wandb-core"}
2
- {"time":"2026-08-06T22:02:10.439399758Z","level":"INFO","msg":"stream: starting","core version":"0.28.1"}
3
- {"time":"2026-08-06T22:02:10.697825711Z","level":"INFO","msg":"stream: created new stream","id":"y74jog3n"}
4
- {"time":"2026-08-06T22:02:10.697928817Z","level":"INFO","msg":"handler: started"}
5
- {"time":"2026-08-06T22:02:10.698047387Z","level":"INFO","msg":"stream: started"}
6
- {"time":"2026-08-06T22:02:10.698057838Z","level":"INFO","msg":"writer: started","stream_id":"y74jog3n"}
7
- {"time":"2026-08-06T22:02:10.698085683Z","level":"INFO","msg":"sender: started"}
8
- {"time":"2026-08-06T22:02:11.245052861Z","level":"INFO","msg":"filestream: sending request","total_files":1,"console_offset":0,"console_lines":1}
9
- {"time":"2026-08-06T22:02:11.348496702Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
10
- {"time":"2026-08-06T22:02:16.001694625Z","level":"INFO","msg":"flowcontrol: backed up, offloading to disk","recordNumber":75}
11
- {"time":"2026-08-06T22:02:16.032767156Z","level":"INFO","msg":"flowcontrol: unblocked","totalOffloaded":9439}
12
- {"time":"2026-08-06T22:02:26.268329691Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":0,"history_lines":3,"events_offset":0,"events_lines":2,"console_offset":1,"console_lines":6,"uploaded_len":2}
13
- {"time":"2026-08-06T22:02:26.831941366Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
14
- {"time":"2026-08-06T22:02:41.257134262Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":3,"history_lines":5,"events_offset":2,"events_lines":2,"console_offset":3,"console_lines":1}
15
- {"time":"2026-08-06T22:02:42.123801353Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
16
- {"time":"2026-08-06T22:02:56.254782559Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":8,"history_lines":2,"events_offset":4,"events_lines":2,"console_offset":3,"console_lines":1}
17
- {"time":"2026-08-06T22:02:57.094993627Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
18
- {"time":"2026-08-06T22:03:11.265458823Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":10,"history_lines":3,"events_offset":6,"events_lines":2,"console_offset":3,"console_lines":1}
19
- {"time":"2026-08-06T22:03:12.009485123Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
20
- {"time":"2026-08-06T22:03:26.260845284Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":13,"history_lines":5,"events_offset":8,"events_lines":2,"console_offset":3,"console_lines":1}
21
- {"time":"2026-08-06T22:03:27.030852793Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
22
- {"time":"2026-08-06T22:03:41.255004725Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":18,"history_lines":3,"events_offset":10,"events_lines":2,"console_offset":3,"console_lines":1}
23
- {"time":"2026-08-06T22:03:42.192468259Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
24
- {"time":"2026-08-06T22:03:56.263062149Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":21,"history_lines":2,"events_offset":12,"events_lines":2,"console_offset":3,"console_lines":1}
25
- {"time":"2026-08-06T22:03:57.052026068Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
26
- {"time":"2026-08-06T22:04:11.254447095Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":23,"history_lines":4,"events_offset":14,"events_lines":2,"console_offset":3,"console_lines":1}
27
- {"time":"2026-08-06T22:04:11.999000035Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
28
- {"time":"2026-08-06T22:04:26.259376734Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":27,"history_lines":5,"events_offset":16,"events_lines":2,"console_offset":7,"console_lines":35}
29
- {"time":"2026-08-06T22:04:26.962933158Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
30
- {"time":"2026-08-06T22:04:41.253315539Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":32,"history_lines":1,"events_offset":18,"events_lines":2,"console_offset":35,"console_lines":1}
31
- {"time":"2026-08-06T22:04:41.939592823Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
32
- {"time":"2026-08-06T22:04:56.273055382Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":33,"history_lines":4,"events_offset":20,"events_lines":2,"console_offset":41,"console_lines":7}
33
- {"time":"2026-08-06T22:04:57.037136274Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
34
- {"time":"2026-08-06T22:05:11.253862511Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":37,"history_lines":4,"events_offset":22,"events_lines":2,"console_offset":43,"console_lines":1}
35
- {"time":"2026-08-06T22:05:12.016225948Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
36
- {"time":"2026-08-06T22:05:26.257933026Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":41,"history_lines":2,"events_offset":24,"events_lines":2,"console_offset":43,"console_lines":1}
37
- {"time":"2026-08-06T22:05:27.024432683Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
38
- {"time":"2026-08-06T22:05:41.26517801Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":43,"history_lines":4,"events_offset":26,"events_lines":2,"console_offset":43,"console_lines":1}
39
- {"time":"2026-08-06T22:05:42.022163128Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
40
- {"time":"2026-08-06T22:05:56.258337796Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":47,"history_lines":4,"events_offset":28,"events_lines":2,"console_offset":43,"console_lines":1}
41
- {"time":"2026-08-06T22:05:56.999137455Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
42
- {"time":"2026-08-06T22:06:11.25661141Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":51,"history_lines":3,"events_offset":30,"events_lines":2,"console_offset":43,"console_lines":1}
43
- {"time":"2026-08-06T22:06:11.994405709Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
44
- {"time":"2026-08-06T22:06:26.264028494Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":54,"history_lines":3,"events_offset":32,"events_lines":2,"console_offset":43,"console_lines":1}
45
- {"time":"2026-08-06T22:06:27.039669461Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
 
1
+ {"time":"2026-08-06T22:10:51.560889469Z","level":"INFO","msg":"wandb-core"}
2
+ {"time":"2026-08-06T22:10:51.561040889Z","level":"INFO","msg":"stream: starting","core version":"0.28.1"}
3
+ {"time":"2026-08-06T22:10:51.824453065Z","level":"INFO","msg":"stream: created new stream","id":"mmhv6fed"}
4
+ {"time":"2026-08-06T22:10:51.824520599Z","level":"INFO","msg":"handler: started"}
5
+ {"time":"2026-08-06T22:10:51.824623415Z","level":"INFO","msg":"stream: started"}
6
+ {"time":"2026-08-06T22:10:51.824635807Z","level":"INFO","msg":"writer: started","stream_id":"mmhv6fed"}
7
+ {"time":"2026-08-06T22:10:51.824649794Z","level":"INFO","msg":"sender: started"}
8
+ {"time":"2026-08-06T22:10:52.410732504Z","level":"INFO","msg":"filestream: sending request","total_files":1,"console_offset":0,"console_lines":1}
9
+ {"time":"2026-08-06T22:10:52.510603182Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
10
+ {"time":"2026-08-06T22:10:57.074452741Z","level":"INFO","msg":"flowcontrol: backed up, offloading to disk","recordNumber":75}
11
+ {"time":"2026-08-06T22:10:57.100284509Z","level":"INFO","msg":"flowcontrol: unblocked","totalOffloaded":9163}
12
+ {"time":"2026-08-06T22:11:07.431056645Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":0,"history_lines":3,"events_offset":0,"events_lines":2,"console_offset":1,"console_lines":6,"uploaded_len":2}
13
+ {"time":"2026-08-06T22:11:07.952818264Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
14
+ {"time":"2026-08-06T22:11:22.422377118Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":3,"history_lines":5,"events_offset":2,"events_lines":2,"console_offset":3,"console_lines":1}
15
+ {"time":"2026-08-06T22:11:23.043122927Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
16
+ {"time":"2026-08-06T22:11:37.420253551Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":8,"history_lines":2,"events_offset":4,"events_lines":2,"console_offset":3,"console_lines":1}
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
wandb/debug.log CHANGED
@@ -1,23 +1,23 @@
1
- 2026-08-06 22:02:10,437 INFO MainThread:1282207 [wandb_setup.py:_flush():81] Current SDK version is 0.28.1
2
- 2026-08-06 22:02:10,437 INFO MainThread:1282207 [wandb_setup.py:_flush():81] Configure stats pid to 1282207
3
- 2026-08-06 22:02:10,437 INFO MainThread:1282207 [wandb_setup.py:_flush():81] Loading settings from environment variables
4
- 2026-08-06 22:02:10,437 INFO MainThread:1282207 [wandb_init.py:setup_run_log_directory():729] Logging user logs to /mnt/data/zainulabideen/zain-exp/notebooks/Activation/wandb/run-20260806_220210-y74jog3n/logs/debug.log
5
- 2026-08-06 22:02:10,437 INFO MainThread:1282207 [wandb_init.py:setup_run_log_directory():730] Logging internal logs to /mnt/data/zainulabideen/zain-exp/notebooks/Activation/wandb/run-20260806_220210-y74jog3n/logs/debug-internal.log
6
- 2026-08-06 22:02:10,437 INFO MainThread:1282207 [wandb_init.py:init():772] calling init triggers
7
- 2026-08-06 22:02:10,438 INFO MainThread:1282207 [wandb_init.py:init():777] wandb.init called with sweep_config: {}
8
  config: {'_wandb': {}}
9
- 2026-08-06 22:02:10,438 INFO MainThread:1282207 [wandb_init.py:init():820] starting backend
10
- 2026-08-06 22:02:10,438 INFO MainThread:1282207 [wandb_init.py:init():826] Connected to an existing wandb-core service via WANDB_SERVICE
11
- 2026-08-06 22:02:10,438 INFO MainThread:1282207 [wandb_init.py:init():835] sending inform_init request
12
- 2026-08-06 22:02:10,698 INFO MainThread:1282207 [wandb_init.py:init():840] backend started and connected
13
- 2026-08-06 22:02:10,701 INFO MainThread:1282207 [wandb_init.py:init():910] updated telemetry
14
- 2026-08-06 22:02:10,708 INFO MainThread:1282207 [wandb_init.py:init():933] communicating run to backend with 90.0 second timeout
15
- 2026-08-06 22:02:10,967 INFO MainThread:1282207 [wandb_init.py:init():978] starting run threads in backend
16
- 2026-08-06 22:02:11,040 INFO MainThread:1282207 [wandb_run.py:_console_start():2621] atexit reg
17
- 2026-08-06 22:02:11,040 INFO MainThread:1282207 [wandb_run.py:_redirect():2471] redirect: wrap_raw
18
- 2026-08-06 22:02:11,040 INFO MainThread:1282207 [wandb_run.py:_redirect():2540] Wrapping output streams.
19
- 2026-08-06 22:02:11,040 INFO MainThread:1282207 [wandb_run.py:_redirect():2563] Redirects installed.
20
- 2026-08-06 22:02:11,043 INFO MainThread:1282207 [wandb_init.py:init():1016] run started, returning control to user process
21
- 2026-08-06 22:02:11,044 INFO MainThread:1282207 [wandb_run.py:_config_callback():1346] config_cb None None {'transformers_version': '5.15.0.dev0', 'architectures': None, 'output_hidden_states': False, 'return_dict': True, 'dtype': None, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'problem_type': None, 'vocab_size': 4096, 'hidden_size': 128, 'intermediate_size': 256, 'num_hidden_layers': 46, 'num_attention_heads': 4, 'num_key_value_heads': 4, 'hidden_act': 'silu', 'max_position_embeddings': 512, 'initializer_range': 0.02, 'rms_norm_eps': 1e-06, 'use_cache': False, 'pad_token_id': 0, 'bos_token_id': 1, 'eos_token_id': 2, 'pretraining_tp': 1, 'tie_word_embeddings': True, 'rope_parameters': {'rope_theta': 10000.0, 'rope_type': 'default'}, 'attention_bias': False, 'attention_dropout': 0.0, 'mlp_bias': False, 'head_dim': 32, '_name_or_path': '', 'tokenizer_name': 'w-ahmad/tiny-stories-tokenizer', 'mlp_type': 'glu', 'activation': 'situglu', 'model_type': 'tiny_llama', 'output_attentions': False, 'output_dir': 'out/glu-situglu-46L_run', 'per_device_train_batch_size': 64, 'num_train_epochs': 1, 'max_steps': 2000, 'learning_rate': 0.0003, 'lr_scheduler_type': 'constant', 'lr_scheduler_kwargs': None, 'warmup_steps': 0, 'optim': 'adamw_torch', 'optim_args': None, 'weight_decay': 0.01, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'optim_target_modules': None, 'gradient_accumulation_steps': 1, 'average_tokens_across_devices': True, 'max_grad_norm': 1.0, 'label_smoothing_factor': 0.0, 'bf16': True, 'fp16': False, 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'gradient_checkpointing': False, 'gradient_checkpointing_kwargs': None, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'use_liger_kernel': False, 'liger_kernel_config': None, 'neftune_noise_alpha': None, 'torch_empty_cache_steps': None, 'auto_find_batch_size': False, 'logging_strategy': 'steps', 'logging_steps': 20, 'logging_first_step': False, 'log_on_each_node': True, 'logging_nan_inf_filter': True, 'include_num_input_tokens_seen': 'no', 'log_level': 'passive', 'log_level_replica': 'warning', 'disable_tqdm': False, 'report_to': ['wandb'], 'run_name': 'LM-glu-situglu-46L-8.1M-20260806-220208', 'project': 'huggingface', 'trackio_space_id': None, 'trackio_bucket_id': None, 'trackio_static_space_id': None, 'eval_strategy': 'steps', 'eval_steps': 200, 'eval_delay': 0, 'per_device_eval_batch_size': 64, 'prediction_loss_only': False, 'eval_on_start': False, 'eval_do_concat_batches': True, 'eval_use_gather_object': False, 'eval_accumulation_steps': None, 'include_for_metrics': [], 'batch_eval_metrics': False, 'save_only_model': False, 'save_strategy': 'steps', 'save_steps': 500, 'save_on_each_node': False, 'save_total_limit': None, 'enable_jit_checkpoint': False, 'push_to_hub': True, 'hub_token': '<HUB_TOKEN>', 'hub_private_repo': None, 'hub_model_id': 'w-ahmad/ACTio-glu-situglu-46L', 'hub_strategy': 'every_save', 'hub_always_push': False, 'hub_revision': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'restore_callback_states_from_checkpoint': False, 'full_determinism': False, 'seed': 42, 'data_seed': 42, 'use_cpu': False, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}, 'parallelism_config': None, 'dataloader_drop_last': False, 'dataloader_num_workers': 0, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'dataloader_prefetch_factor': None, 'dataloader_multiprocessing_context': None, 'dataloader_in_order': True, 'remove_unused_columns': False, 'label_names': None, 'train_sampling_strategy': 'random', 'length_column_name': 'length', 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'ddp_static_graph': None, 'ddp_backend': None, 'ddp_timeout': 1800, 'fsdp': None, 'fsdp_config': None, 'deepspeed': None, 'debug': [], 'skip_memory_metrics': True, 'do_train': False, 'do_eval': True, 'do_predict': False, 'resume_from_checkpoint': None, 'local_rank': -1}
22
- 2026-08-06 22:02:11,047 INFO MainThread:1282207 [wandb_config.py:__setitem__():155] [no run ID] config set model/num_parameters = 8072832 - <bound method Run._config_callback of <wandb.sdk.wandb_run.Run object at 0x147585083210>>
23
- 2026-08-06 22:02:11,047 INFO MainThread:1282207 [wandb_run.py:_config_callback():1346] config_cb model/num_parameters 8072832 None
 
1
+ 2026-08-06 22:10:51,557 INFO MainThread:1365488 [wandb_setup.py:_flush():81] Current SDK version is 0.28.1
2
+ 2026-08-06 22:10:51,559 INFO MainThread:1365488 [wandb_setup.py:_flush():81] Configure stats pid to 1365488
3
+ 2026-08-06 22:10:51,559 INFO MainThread:1365488 [wandb_setup.py:_flush():81] Loading settings from environment variables
4
+ 2026-08-06 22:10:51,559 INFO MainThread:1365488 [wandb_init.py:setup_run_log_directory():729] Logging user logs to /mnt/data/zainulabideen/zain-exp/notebooks/Activation/wandb/run-20260806_221051-mmhv6fed/logs/debug.log
5
+ 2026-08-06 22:10:51,559 INFO MainThread:1365488 [wandb_init.py:setup_run_log_directory():730] Logging internal logs to /mnt/data/zainulabideen/zain-exp/notebooks/Activation/wandb/run-20260806_221051-mmhv6fed/logs/debug-internal.log
6
+ 2026-08-06 22:10:51,559 INFO MainThread:1365488 [wandb_init.py:init():772] calling init triggers
7
+ 2026-08-06 22:10:51,559 INFO MainThread:1365488 [wandb_init.py:init():777] wandb.init called with sweep_config: {}
8
  config: {'_wandb': {}}
9
+ 2026-08-06 22:10:51,559 INFO MainThread:1365488 [wandb_init.py:init():820] starting backend
10
+ 2026-08-06 22:10:51,559 INFO MainThread:1365488 [wandb_init.py:init():826] Connected to an existing wandb-core service via WANDB_SERVICE
11
+ 2026-08-06 22:10:51,559 INFO MainThread:1365488 [wandb_init.py:init():835] sending inform_init request
12
+ 2026-08-06 22:10:51,825 INFO MainThread:1365488 [wandb_init.py:init():840] backend started and connected
13
+ 2026-08-06 22:10:51,828 INFO MainThread:1365488 [wandb_init.py:init():910] updated telemetry
14
+ 2026-08-06 22:10:51,835 INFO MainThread:1365488 [wandb_init.py:init():933] communicating run to backend with 90.0 second timeout
15
+ 2026-08-06 22:10:52,146 INFO MainThread:1365488 [wandb_init.py:init():978] starting run threads in backend
16
+ 2026-08-06 22:10:52,218 INFO MainThread:1365488 [wandb_run.py:_console_start():2621] atexit reg
17
+ 2026-08-06 22:10:52,218 INFO MainThread:1365488 [wandb_run.py:_redirect():2471] redirect: wrap_raw
18
+ 2026-08-06 22:10:52,218 INFO MainThread:1365488 [wandb_run.py:_redirect():2540] Wrapping output streams.
19
+ 2026-08-06 22:10:52,219 INFO MainThread:1365488 [wandb_run.py:_redirect():2563] Redirects installed.
20
+ 2026-08-06 22:10:52,222 INFO MainThread:1365488 [wandb_init.py:init():1016] run started, returning control to user process
21
+ 2026-08-06 22:10:52,223 INFO MainThread:1365488 [wandb_run.py:_config_callback():1346] config_cb None None {'transformers_version': '5.15.0.dev0', 'architectures': None, 'output_hidden_states': False, 'return_dict': True, 'dtype': None, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'problem_type': None, 'vocab_size': 4096, 'hidden_size': 128, 'intermediate_size': 256, 'num_hidden_layers': 46, 'num_attention_heads': 4, 'num_key_value_heads': 4, 'hidden_act': 'silu', 'max_position_embeddings': 512, 'initializer_range': 0.02, 'rms_norm_eps': 1e-06, 'use_cache': False, 'pad_token_id': 0, 'bos_token_id': 1, 'eos_token_id': 2, 'pretraining_tp': 1, 'tie_word_embeddings': True, 'rope_parameters': {'rope_theta': 10000.0, 'rope_type': 'default'}, 'attention_bias': False, 'attention_dropout': 0.0, 'mlp_bias': False, 'head_dim': 32, '_name_or_path': '', 'tokenizer_name': 'w-ahmad/tiny-stories-tokenizer', 'mlp_type': 'glu', 'activation': 's10', 'model_type': 'tiny_llama', 'output_attentions': False, 'output_dir': 'out/glu-s10-46L_run', 'per_device_train_batch_size': 64, 'num_train_epochs': 1, 'max_steps': 2000, 'learning_rate': 0.0003, 'lr_scheduler_type': 'constant', 'lr_scheduler_kwargs': None, 'warmup_steps': 0, 'optim': 'adamw_torch', 'optim_args': None, 'weight_decay': 0.01, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'optim_target_modules': None, 'gradient_accumulation_steps': 1, 'average_tokens_across_devices': True, 'max_grad_norm': 1.0, 'label_smoothing_factor': 0.0, 'bf16': True, 'fp16': False, 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'gradient_checkpointing': False, 'gradient_checkpointing_kwargs': None, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'use_liger_kernel': False, 'liger_kernel_config': None, 'neftune_noise_alpha': None, 'torch_empty_cache_steps': None, 'auto_find_batch_size': False, 'logging_strategy': 'steps', 'logging_steps': 20, 'logging_first_step': False, 'log_on_each_node': True, 'logging_nan_inf_filter': True, 'include_num_input_tokens_seen': 'no', 'log_level': 'passive', 'log_level_replica': 'warning', 'disable_tqdm': False, 'report_to': ['wandb'], 'run_name': 'LM-glu-s10-46L-8.1M-20260806-221050', 'project': 'huggingface', 'trackio_space_id': None, 'trackio_bucket_id': None, 'trackio_static_space_id': None, 'eval_strategy': 'steps', 'eval_steps': 200, 'eval_delay': 0, 'per_device_eval_batch_size': 64, 'prediction_loss_only': False, 'eval_on_start': False, 'eval_do_concat_batches': True, 'eval_use_gather_object': False, 'eval_accumulation_steps': None, 'include_for_metrics': [], 'batch_eval_metrics': False, 'save_only_model': False, 'save_strategy': 'steps', 'save_steps': 500, 'save_on_each_node': False, 'save_total_limit': None, 'enable_jit_checkpoint': False, 'push_to_hub': True, 'hub_token': '<HUB_TOKEN>', 'hub_private_repo': None, 'hub_model_id': 'w-ahmad/ACTio-glu-s10-46L', 'hub_strategy': 'every_save', 'hub_always_push': False, 'hub_revision': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'restore_callback_states_from_checkpoint': False, 'full_determinism': False, 'seed': 42, 'data_seed': 42, 'use_cpu': False, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}, 'parallelism_config': None, 'dataloader_drop_last': False, 'dataloader_num_workers': 0, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'dataloader_prefetch_factor': None, 'dataloader_multiprocessing_context': None, 'dataloader_in_order': True, 'remove_unused_columns': False, 'label_names': None, 'train_sampling_strategy': 'random', 'length_column_name': 'length', 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'ddp_static_graph': None, 'ddp_backend': None, 'ddp_timeout': 1800, 'fsdp': None, 'fsdp_config': None, 'deepspeed': None, 'debug': [], 'skip_memory_metrics': True, 'do_train': False, 'do_eval': True, 'do_predict': False, 'resume_from_checkpoint': None, 'local_rank': -1}
22
+ 2026-08-06 22:10:52,226 INFO MainThread:1365488 [wandb_config.py:__setitem__():155] [no run ID] config set model/num_parameters = 8072832 - <bound method Run._config_callback of <wandb.sdk.wandb_run.Run object at 0x1499e0521dd0>>
23
+ 2026-08-06 22:10:52,226 INFO MainThread:1365488 [wandb_run.py:_config_callback():1346] config_cb model/num_parameters 8072832 None
wandb/run-20260806_202910-50qq810b/logs/debug-core.log CHANGED
@@ -294,3 +294,15 @@
294
  {"time":"2026-08-06T22:02:10.439113476Z","level":"INFO","msg":"handleInformInit: received","streamId":"y74jog3n","id":"19(@)"}
295
  {"time":"2026-08-06T22:02:10.698054902Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"y74jog3n","id":"19(@)"}
296
  {"time":"2026-08-06T22:02:16.042923531Z","level":"INFO","msg":"connection: cancelling request","id":"19(@)","requestId":"zee4komq248b"}
 
 
 
 
 
 
 
 
 
 
 
 
 
294
  {"time":"2026-08-06T22:02:10.439113476Z","level":"INFO","msg":"handleInformInit: received","streamId":"y74jog3n","id":"19(@)"}
295
  {"time":"2026-08-06T22:02:10.698054902Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"y74jog3n","id":"19(@)"}
296
  {"time":"2026-08-06T22:02:16.042923531Z","level":"INFO","msg":"connection: cancelling request","id":"19(@)","requestId":"zee4komq248b"}
297
+ {"time":"2026-08-06T22:10:40.201957301Z","level":"INFO","msg":"connection: cancelling request","id":"19(@)","requestId":"zee4komq248b"}
298
+ {"time":"2026-08-06T22:10:41.673521316Z","level":"INFO","msg":"connection: cancelling request","id":"19(@)","requestId":"zee4komq248b"}
299
+ {"time":"2026-08-06T22:10:41.690938257Z","level":"INFO","msg":"handleInformFinish: finish message received","streamId":"y74jog3n","id":"19(@)"}
300
+ {"time":"2026-08-06T22:10:41.691692107Z","level":"INFO","msg":"handleInformFinish: stream closed","streamId":"y74jog3n","id":"19(@)"}
301
+ {"time":"2026-08-06T22:10:43.724610654Z","level":"INFO","msg":"connection: closing","id":"19(@)"}
302
+ {"time":"2026-08-06T22:10:43.724616012Z","level":"INFO","msg":"processOutgoingData: finished","id":"19(@)"}
303
+ {"time":"2026-08-06T22:10:43.724709021Z","level":"INFO","msg":"connection: closed successfully","id":"19(@)"}
304
+ {"time":"2026-08-06T22:10:43.724714942Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"19(@)"}
305
+ {"time":"2026-08-06T22:10:51.347033931Z","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"20(@)"}
306
+ {"time":"2026-08-06T22:10:51.560725873Z","level":"INFO","msg":"handleInformInit: received","streamId":"mmhv6fed","id":"20(@)"}
307
+ {"time":"2026-08-06T22:10:51.824628616Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"mmhv6fed","id":"20(@)"}
308
+ {"time":"2026-08-06T22:10:57.222474886Z","level":"INFO","msg":"connection: cancelling request","id":"20(@)","requestId":"m8oneq7o9rzj"}