yihanwang617 commited on
Commit
90e70b4
1 Parent(s): 2ce31a9

Model save

Browse files
README.md ADDED
@@ -0,0 +1,65 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T
4
+ tags:
5
+ - trl
6
+ - sft
7
+ - generated_from_trainer
8
+ model-index:
9
+ - name: tinyllama-sft-vicuna-random-100k
10
+ results: []
11
+ ---
12
+
13
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
14
+ should probably proofread and complete it, then remove this comment. -->
15
+
16
+ # tinyllama-sft-vicuna-random-100k
17
+
18
+ This model is a fine-tuned version of [TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T](https://huggingface.co/TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T) on the None dataset.
19
+ It achieves the following results on the evaluation set:
20
+ - Loss: 0.7502
21
+
22
+ ## Model description
23
+
24
+ More information needed
25
+
26
+ ## Intended uses & limitations
27
+
28
+ More information needed
29
+
30
+ ## Training and evaluation data
31
+
32
+ More information needed
33
+
34
+ ## Training procedure
35
+
36
+ ### Training hyperparameters
37
+
38
+ The following hyperparameters were used during training:
39
+ - learning_rate: 2e-05
40
+ - train_batch_size: 16
41
+ - eval_batch_size: 8
42
+ - seed: 42
43
+ - distributed_type: multi-GPU
44
+ - num_devices: 4
45
+ - gradient_accumulation_steps: 2
46
+ - total_train_batch_size: 128
47
+ - total_eval_batch_size: 32
48
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
49
+ - lr_scheduler_type: cosine
50
+ - lr_scheduler_warmup_ratio: 0.1
51
+ - num_epochs: 1
52
+
53
+ ### Training results
54
+
55
+ | Training Loss | Epoch | Step | Validation Loss |
56
+ |:-------------:|:-----:|:----:|:---------------:|
57
+ | 0.7072 | 1.0 | 703 | 0.7502 |
58
+
59
+
60
+ ### Framework versions
61
+
62
+ - Transformers 4.39.0.dev0
63
+ - Pytorch 2.2.1+cu121
64
+ - Datasets 2.14.6
65
+ - Tokenizers 0.15.0
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.7627222812701425,
4
+ "train_runtime": 5862.0034,
5
+ "train_samples": 90000,
6
+ "train_samples_per_second": 15.353,
7
+ "train_steps_per_second": 0.12
8
+ }
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 1,
3
+ "eos_token_id": 2,
4
+ "max_length": 2048,
5
+ "pad_token_id": 0,
6
+ "transformers_version": "4.39.0.dev0"
7
+ }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:56477c25ebcdf018c2daecedf2921be6c1e8316a107468229806712339b24bed
3
  size 2200128056
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:def68e8808c97e039dee5dba5be22e198c18b820a53c8f11f14d0927bdfd0e9c
3
  size 2200128056
runs/May23_00-22-04_nova.cs.ucla.edu/events.out.tfevents.1716448938.nova.cs.ucla.edu.3094888.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:94d4d34d075a14b9daae14656d88dc364ee927890a11b4a7b2d452ecdb552c3d
3
- size 34514
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b0a1d2d95d4956a691ab59d3f18bf8db0458818d74a297b5e3eb4733f3230d84
3
+ size 35139
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.7627222812701425,
4
+ "train_runtime": 5862.0034,
5
+ "train_samples": 90000,
6
+ "train_samples_per_second": 15.353,
7
+ "train_steps_per_second": 0.12
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1025 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9992892679459844,
5
+ "eval_steps": 500,
6
+ "global_step": 703,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "grad_norm": 8.976748246707771,
14
+ "learning_rate": 2.8169014084507043e-07,
15
+ "loss": 1.2046,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.01,
20
+ "grad_norm": 9.141400774480953,
21
+ "learning_rate": 1.4084507042253523e-06,
22
+ "loss": 1.212,
23
+ "step": 5
24
+ },
25
+ {
26
+ "epoch": 0.01,
27
+ "grad_norm": 9.49864441857399,
28
+ "learning_rate": 2.8169014084507046e-06,
29
+ "loss": 1.1487,
30
+ "step": 10
31
+ },
32
+ {
33
+ "epoch": 0.02,
34
+ "grad_norm": 3.2926935702992415,
35
+ "learning_rate": 4.225352112676057e-06,
36
+ "loss": 1.0049,
37
+ "step": 15
38
+ },
39
+ {
40
+ "epoch": 0.03,
41
+ "grad_norm": 1.7117549060658137,
42
+ "learning_rate": 5.633802816901409e-06,
43
+ "loss": 0.9194,
44
+ "step": 20
45
+ },
46
+ {
47
+ "epoch": 0.04,
48
+ "grad_norm": 1.2735260163319009,
49
+ "learning_rate": 7.042253521126761e-06,
50
+ "loss": 0.8716,
51
+ "step": 25
52
+ },
53
+ {
54
+ "epoch": 0.04,
55
+ "grad_norm": 1.0790301127605728,
56
+ "learning_rate": 8.450704225352114e-06,
57
+ "loss": 0.864,
58
+ "step": 30
59
+ },
60
+ {
61
+ "epoch": 0.05,
62
+ "grad_norm": 0.7414273564854603,
63
+ "learning_rate": 9.859154929577466e-06,
64
+ "loss": 0.8359,
65
+ "step": 35
66
+ },
67
+ {
68
+ "epoch": 0.06,
69
+ "grad_norm": 0.6656956199139998,
70
+ "learning_rate": 1.1267605633802819e-05,
71
+ "loss": 0.8614,
72
+ "step": 40
73
+ },
74
+ {
75
+ "epoch": 0.06,
76
+ "grad_norm": 0.6590669608248467,
77
+ "learning_rate": 1.2676056338028171e-05,
78
+ "loss": 0.8375,
79
+ "step": 45
80
+ },
81
+ {
82
+ "epoch": 0.07,
83
+ "grad_norm": 0.6842437677529626,
84
+ "learning_rate": 1.4084507042253522e-05,
85
+ "loss": 0.8306,
86
+ "step": 50
87
+ },
88
+ {
89
+ "epoch": 0.08,
90
+ "grad_norm": 0.6080481338431981,
91
+ "learning_rate": 1.5492957746478872e-05,
92
+ "loss": 0.8391,
93
+ "step": 55
94
+ },
95
+ {
96
+ "epoch": 0.09,
97
+ "grad_norm": 0.5829909243761902,
98
+ "learning_rate": 1.6901408450704228e-05,
99
+ "loss": 0.8211,
100
+ "step": 60
101
+ },
102
+ {
103
+ "epoch": 0.09,
104
+ "grad_norm": 0.5774288402112024,
105
+ "learning_rate": 1.830985915492958e-05,
106
+ "loss": 0.8104,
107
+ "step": 65
108
+ },
109
+ {
110
+ "epoch": 0.1,
111
+ "grad_norm": 0.6037500273154381,
112
+ "learning_rate": 1.9718309859154933e-05,
113
+ "loss": 0.7882,
114
+ "step": 70
115
+ },
116
+ {
117
+ "epoch": 0.11,
118
+ "grad_norm": 0.5823339744513759,
119
+ "learning_rate": 1.9998023297700656e-05,
120
+ "loss": 0.8226,
121
+ "step": 75
122
+ },
123
+ {
124
+ "epoch": 0.11,
125
+ "grad_norm": 0.5970199949943188,
126
+ "learning_rate": 1.9989994283927287e-05,
127
+ "loss": 0.7805,
128
+ "step": 80
129
+ },
130
+ {
131
+ "epoch": 0.12,
132
+ "grad_norm": 0.5863046182186602,
133
+ "learning_rate": 1.997579437055642e-05,
134
+ "loss": 0.8011,
135
+ "step": 85
136
+ },
137
+ {
138
+ "epoch": 0.13,
139
+ "grad_norm": 0.6404658915624578,
140
+ "learning_rate": 1.9955432328988437e-05,
141
+ "loss": 0.8092,
142
+ "step": 90
143
+ },
144
+ {
145
+ "epoch": 0.14,
146
+ "grad_norm": 0.5889469279145607,
147
+ "learning_rate": 1.9928920737019735e-05,
148
+ "loss": 0.8111,
149
+ "step": 95
150
+ },
151
+ {
152
+ "epoch": 0.14,
153
+ "grad_norm": 0.5999505405000287,
154
+ "learning_rate": 1.9896275971073326e-05,
155
+ "loss": 0.8034,
156
+ "step": 100
157
+ },
158
+ {
159
+ "epoch": 0.15,
160
+ "grad_norm": 0.5958697553643331,
161
+ "learning_rate": 1.9857518196082964e-05,
162
+ "loss": 0.783,
163
+ "step": 105
164
+ },
165
+ {
166
+ "epoch": 0.16,
167
+ "grad_norm": 0.6235891380868392,
168
+ "learning_rate": 1.981267135303714e-05,
169
+ "loss": 0.7986,
170
+ "step": 110
171
+ },
172
+ {
173
+ "epoch": 0.16,
174
+ "grad_norm": 0.6386697510622403,
175
+ "learning_rate": 1.976176314419051e-05,
176
+ "loss": 0.8173,
177
+ "step": 115
178
+ },
179
+ {
180
+ "epoch": 0.17,
181
+ "grad_norm": 0.6223044203965952,
182
+ "learning_rate": 1.9704825015952005e-05,
183
+ "loss": 0.7981,
184
+ "step": 120
185
+ },
186
+ {
187
+ "epoch": 0.18,
188
+ "grad_norm": 0.6801656349332326,
189
+ "learning_rate": 1.9641892139460133e-05,
190
+ "loss": 0.8018,
191
+ "step": 125
192
+ },
193
+ {
194
+ "epoch": 0.18,
195
+ "grad_norm": 0.5736017682468991,
196
+ "learning_rate": 1.9573003388857476e-05,
197
+ "loss": 0.7801,
198
+ "step": 130
199
+ },
200
+ {
201
+ "epoch": 0.19,
202
+ "grad_norm": 0.6377284634315731,
203
+ "learning_rate": 1.949820131727783e-05,
204
+ "loss": 0.8233,
205
+ "step": 135
206
+ },
207
+ {
208
+ "epoch": 0.2,
209
+ "grad_norm": 0.5834471085873422,
210
+ "learning_rate": 1.9417532130560784e-05,
211
+ "loss": 0.7792,
212
+ "step": 140
213
+ },
214
+ {
215
+ "epoch": 0.21,
216
+ "grad_norm": 0.5750039228232332,
217
+ "learning_rate": 1.933104565871001e-05,
218
+ "loss": 0.79,
219
+ "step": 145
220
+ },
221
+ {
222
+ "epoch": 0.21,
223
+ "grad_norm": 0.5788157023616078,
224
+ "learning_rate": 1.9238795325112867e-05,
225
+ "loss": 0.7823,
226
+ "step": 150
227
+ },
228
+ {
229
+ "epoch": 0.22,
230
+ "grad_norm": 0.5905988672413683,
231
+ "learning_rate": 1.9140838113540347e-05,
232
+ "loss": 0.7907,
233
+ "step": 155
234
+ },
235
+ {
236
+ "epoch": 0.23,
237
+ "grad_norm": 0.6036919615732501,
238
+ "learning_rate": 1.9037234532947737e-05,
239
+ "loss": 0.7859,
240
+ "step": 160
241
+ },
242
+ {
243
+ "epoch": 0.23,
244
+ "grad_norm": 0.5744116424871404,
245
+ "learning_rate": 1.8928048580097758e-05,
246
+ "loss": 0.7902,
247
+ "step": 165
248
+ },
249
+ {
250
+ "epoch": 0.24,
251
+ "grad_norm": 0.5858290882445865,
252
+ "learning_rate": 1.8813347700029244e-05,
253
+ "loss": 0.7756,
254
+ "step": 170
255
+ },
256
+ {
257
+ "epoch": 0.25,
258
+ "grad_norm": 0.6203103361508707,
259
+ "learning_rate": 1.869320274439583e-05,
260
+ "loss": 0.7945,
261
+ "step": 175
262
+ },
263
+ {
264
+ "epoch": 0.26,
265
+ "grad_norm": 0.6094752243332237,
266
+ "learning_rate": 1.8567687927700255e-05,
267
+ "loss": 0.8089,
268
+ "step": 180
269
+ },
270
+ {
271
+ "epoch": 0.26,
272
+ "grad_norm": 0.6065177506682852,
273
+ "learning_rate": 1.8436880781451545e-05,
274
+ "loss": 0.7866,
275
+ "step": 185
276
+ },
277
+ {
278
+ "epoch": 0.27,
279
+ "grad_norm": 0.6422514480326399,
280
+ "learning_rate": 1.8300862106273113e-05,
281
+ "loss": 0.7661,
282
+ "step": 190
283
+ },
284
+ {
285
+ "epoch": 0.28,
286
+ "grad_norm": 0.5976731025044114,
287
+ "learning_rate": 1.8159715921991612e-05,
288
+ "loss": 0.7518,
289
+ "step": 195
290
+ },
291
+ {
292
+ "epoch": 0.28,
293
+ "grad_norm": 0.5813547875268166,
294
+ "learning_rate": 1.801352941573718e-05,
295
+ "loss": 0.777,
296
+ "step": 200
297
+ },
298
+ {
299
+ "epoch": 0.29,
300
+ "grad_norm": 0.6079636229603744,
301
+ "learning_rate": 1.786239288808727e-05,
302
+ "loss": 0.7517,
303
+ "step": 205
304
+ },
305
+ {
306
+ "epoch": 0.3,
307
+ "grad_norm": 0.6116028297485965,
308
+ "learning_rate": 1.770639969728726e-05,
309
+ "loss": 0.7582,
310
+ "step": 210
311
+ },
312
+ {
313
+ "epoch": 0.31,
314
+ "grad_norm": 0.6104535774475512,
315
+ "learning_rate": 1.7545646201582304e-05,
316
+ "loss": 0.7739,
317
+ "step": 215
318
+ },
319
+ {
320
+ "epoch": 0.31,
321
+ "grad_norm": 0.6092743685177864,
322
+ "learning_rate": 1.738023169969608e-05,
323
+ "loss": 0.7747,
324
+ "step": 220
325
+ },
326
+ {
327
+ "epoch": 0.32,
328
+ "grad_norm": 0.5879141856316753,
329
+ "learning_rate": 1.721025836949317e-05,
330
+ "loss": 0.7601,
331
+ "step": 225
332
+ },
333
+ {
334
+ "epoch": 0.33,
335
+ "grad_norm": 0.5801521678476529,
336
+ "learning_rate": 1.703583120486297e-05,
337
+ "loss": 0.7831,
338
+ "step": 230
339
+ },
340
+ {
341
+ "epoch": 0.33,
342
+ "grad_norm": 0.6538515323321731,
343
+ "learning_rate": 1.6857057950864134e-05,
344
+ "loss": 0.7792,
345
+ "step": 235
346
+ },
347
+ {
348
+ "epoch": 0.34,
349
+ "grad_norm": 0.6283544032704685,
350
+ "learning_rate": 1.6674049037169565e-05,
351
+ "loss": 0.7699,
352
+ "step": 240
353
+ },
354
+ {
355
+ "epoch": 0.35,
356
+ "grad_norm": 0.5680653535007143,
357
+ "learning_rate": 1.648691750985314e-05,
358
+ "loss": 0.7465,
359
+ "step": 245
360
+ },
361
+ {
362
+ "epoch": 0.36,
363
+ "grad_norm": 0.6108554255786504,
364
+ "learning_rate": 1.6295778961560242e-05,
365
+ "loss": 0.7615,
366
+ "step": 250
367
+ },
368
+ {
369
+ "epoch": 0.36,
370
+ "grad_norm": 0.5595610558009545,
371
+ "learning_rate": 1.6100751460105244e-05,
372
+ "loss": 0.7517,
373
+ "step": 255
374
+ },
375
+ {
376
+ "epoch": 0.37,
377
+ "grad_norm": 0.5638534604629722,
378
+ "learning_rate": 1.5901955475540087e-05,
379
+ "loss": 0.7433,
380
+ "step": 260
381
+ },
382
+ {
383
+ "epoch": 0.38,
384
+ "grad_norm": 0.6290842137244347,
385
+ "learning_rate": 1.5699513805738942e-05,
386
+ "loss": 0.7546,
387
+ "step": 265
388
+ },
389
+ {
390
+ "epoch": 0.38,
391
+ "grad_norm": 0.5899201877158341,
392
+ "learning_rate": 1.549355150054501e-05,
393
+ "loss": 0.7586,
394
+ "step": 270
395
+ },
396
+ {
397
+ "epoch": 0.39,
398
+ "grad_norm": 0.5848454901161203,
399
+ "learning_rate": 1.5284195784526196e-05,
400
+ "loss": 0.7435,
401
+ "step": 275
402
+ },
403
+ {
404
+ "epoch": 0.4,
405
+ "grad_norm": 0.6441906832873957,
406
+ "learning_rate": 1.5071575978387505e-05,
407
+ "loss": 0.7557,
408
+ "step": 280
409
+ },
410
+ {
411
+ "epoch": 0.41,
412
+ "grad_norm": 0.5889076952992223,
413
+ "learning_rate": 1.4855823419088576e-05,
414
+ "loss": 0.7523,
415
+ "step": 285
416
+ },
417
+ {
418
+ "epoch": 0.41,
419
+ "grad_norm": 0.6138750994803515,
420
+ "learning_rate": 1.4637071378715807e-05,
421
+ "loss": 0.7466,
422
+ "step": 290
423
+ },
424
+ {
425
+ "epoch": 0.42,
426
+ "grad_norm": 0.5460450262431842,
427
+ "learning_rate": 1.4415454982159121e-05,
428
+ "loss": 0.7575,
429
+ "step": 295
430
+ },
431
+ {
432
+ "epoch": 0.43,
433
+ "grad_norm": 0.5958584774428966,
434
+ "learning_rate": 1.419111112364422e-05,
435
+ "loss": 0.7651,
436
+ "step": 300
437
+ },
438
+ {
439
+ "epoch": 0.43,
440
+ "grad_norm": 0.5892595546498745,
441
+ "learning_rate": 1.3964178382171942e-05,
442
+ "loss": 0.7509,
443
+ "step": 305
444
+ },
445
+ {
446
+ "epoch": 0.44,
447
+ "grad_norm": 0.5824299769787271,
448
+ "learning_rate": 1.3734796935916888e-05,
449
+ "loss": 0.7248,
450
+ "step": 310
451
+ },
452
+ {
453
+ "epoch": 0.45,
454
+ "grad_norm": 0.564819905580984,
455
+ "learning_rate": 1.3503108475638244e-05,
456
+ "loss": 0.7288,
457
+ "step": 315
458
+ },
459
+ {
460
+ "epoch": 0.45,
461
+ "grad_norm": 0.6263083055380363,
462
+ "learning_rate": 1.326925611715627e-05,
463
+ "loss": 0.7559,
464
+ "step": 320
465
+ },
466
+ {
467
+ "epoch": 0.46,
468
+ "grad_norm": 0.5885749596082486,
469
+ "learning_rate": 1.3033384312948487e-05,
470
+ "loss": 0.7448,
471
+ "step": 325
472
+ },
473
+ {
474
+ "epoch": 0.47,
475
+ "grad_norm": 0.6217412195838338,
476
+ "learning_rate": 1.2795638762920254e-05,
477
+ "loss": 0.742,
478
+ "step": 330
479
+ },
480
+ {
481
+ "epoch": 0.48,
482
+ "grad_norm": 0.6379002045294617,
483
+ "learning_rate": 1.2556166324404747e-05,
484
+ "loss": 0.7099,
485
+ "step": 335
486
+ },
487
+ {
488
+ "epoch": 0.48,
489
+ "grad_norm": 0.5539140087654301,
490
+ "learning_rate": 1.2315114921448012e-05,
491
+ "loss": 0.7362,
492
+ "step": 340
493
+ },
494
+ {
495
+ "epoch": 0.49,
496
+ "grad_norm": 0.5709158729850722,
497
+ "learning_rate": 1.2072633453435092e-05,
498
+ "loss": 0.7259,
499
+ "step": 345
500
+ },
501
+ {
502
+ "epoch": 0.5,
503
+ "grad_norm": 0.6103140402089173,
504
+ "learning_rate": 1.1828871703113686e-05,
505
+ "loss": 0.7525,
506
+ "step": 350
507
+ },
508
+ {
509
+ "epoch": 0.5,
510
+ "grad_norm": 0.6006237395857831,
511
+ "learning_rate": 1.158398024407215e-05,
512
+ "loss": 0.7443,
513
+ "step": 355
514
+ },
515
+ {
516
+ "epoch": 0.51,
517
+ "grad_norm": 0.5730036044160878,
518
+ "learning_rate": 1.1338110347728973e-05,
519
+ "loss": 0.7488,
520
+ "step": 360
521
+ },
522
+ {
523
+ "epoch": 0.52,
524
+ "grad_norm": 0.6357913417816614,
525
+ "learning_rate": 1.1091413889891211e-05,
526
+ "loss": 0.7451,
527
+ "step": 365
528
+ },
529
+ {
530
+ "epoch": 0.53,
531
+ "grad_norm": 0.6484616619442031,
532
+ "learning_rate": 1.0844043256939585e-05,
533
+ "loss": 0.7513,
534
+ "step": 370
535
+ },
536
+ {
537
+ "epoch": 0.53,
538
+ "grad_norm": 0.6147553788096094,
539
+ "learning_rate": 1.05961512516982e-05,
540
+ "loss": 0.7547,
541
+ "step": 375
542
+ },
543
+ {
544
+ "epoch": 0.54,
545
+ "grad_norm": 0.6026348227570094,
546
+ "learning_rate": 1.0347890999046998e-05,
547
+ "loss": 0.7291,
548
+ "step": 380
549
+ },
550
+ {
551
+ "epoch": 0.55,
552
+ "grad_norm": 0.5607221603847458,
553
+ "learning_rate": 1.00994158513353e-05,
554
+ "loss": 0.729,
555
+ "step": 385
556
+ },
557
+ {
558
+ "epoch": 0.55,
559
+ "grad_norm": 0.5908644383211163,
560
+ "learning_rate": 9.850879293654829e-06,
561
+ "loss": 0.7319,
562
+ "step": 390
563
+ },
564
+ {
565
+ "epoch": 0.56,
566
+ "grad_norm": 0.5697397462576617,
567
+ "learning_rate": 9.602434849030747e-06,
568
+ "loss": 0.7522,
569
+ "step": 395
570
+ },
571
+ {
572
+ "epoch": 0.57,
573
+ "grad_norm": 0.5939805294573912,
574
+ "learning_rate": 9.354235983589229e-06,
575
+ "loss": 0.7285,
576
+ "step": 400
577
+ },
578
+ {
579
+ "epoch": 0.58,
580
+ "grad_norm": 0.5755564582128363,
581
+ "learning_rate": 9.106436011760229e-06,
582
+ "loss": 0.7102,
583
+ "step": 405
584
+ },
585
+ {
586
+ "epoch": 0.58,
587
+ "grad_norm": 0.6451432716799574,
588
+ "learning_rate": 8.859188001573916e-06,
589
+ "loss": 0.7511,
590
+ "step": 410
591
+ },
592
+ {
593
+ "epoch": 0.59,
594
+ "grad_norm": 0.5843168824282191,
595
+ "learning_rate": 8.61264468010932e-06,
596
+ "loss": 0.7022,
597
+ "step": 415
598
+ },
599
+ {
600
+ "epoch": 0.6,
601
+ "grad_norm": 0.6112413168467156,
602
+ "learning_rate": 8.3669583391536e-06,
603
+ "loss": 0.7275,
604
+ "step": 420
605
+ },
606
+ {
607
+ "epoch": 0.6,
608
+ "grad_norm": 0.581022238822059,
609
+ "learning_rate": 8.122280741130177e-06,
610
+ "loss": 0.737,
611
+ "step": 425
612
+ },
613
+ {
614
+ "epoch": 0.61,
615
+ "grad_norm": 0.5858683287529846,
616
+ "learning_rate": 7.878763025353875e-06,
617
+ "loss": 0.7456,
618
+ "step": 430
619
+ },
620
+ {
621
+ "epoch": 0.62,
622
+ "grad_norm": 0.5983662476814803,
623
+ "learning_rate": 7.636555614670953e-06,
624
+ "loss": 0.7443,
625
+ "step": 435
626
+ },
627
+ {
628
+ "epoch": 0.63,
629
+ "grad_norm": 0.5853808275325328,
630
+ "learning_rate": 7.395808122541697e-06,
631
+ "loss": 0.7456,
632
+ "step": 440
633
+ },
634
+ {
635
+ "epoch": 0.63,
636
+ "grad_norm": 0.6178373733573391,
637
+ "learning_rate": 7.156669260622997e-06,
638
+ "loss": 0.7272,
639
+ "step": 445
640
+ },
641
+ {
642
+ "epoch": 0.64,
643
+ "grad_norm": 0.5767484461478896,
644
+ "learning_rate": 6.9192867469079625e-06,
645
+ "loss": 0.7355,
646
+ "step": 450
647
+ },
648
+ {
649
+ "epoch": 0.65,
650
+ "grad_norm": 0.6095308999494778,
651
+ "learning_rate": 6.683807214479323e-06,
652
+ "loss": 0.737,
653
+ "step": 455
654
+ },
655
+ {
656
+ "epoch": 0.65,
657
+ "grad_norm": 0.5757726348096623,
658
+ "learning_rate": 6.450376120933008e-06,
659
+ "loss": 0.725,
660
+ "step": 460
661
+ },
662
+ {
663
+ "epoch": 0.66,
664
+ "grad_norm": 0.6084178997751468,
665
+ "learning_rate": 6.219137658527819e-06,
666
+ "loss": 0.7484,
667
+ "step": 465
668
+ },
669
+ {
670
+ "epoch": 0.67,
671
+ "grad_norm": 0.5625041815811782,
672
+ "learning_rate": 5.990234665116713e-06,
673
+ "loss": 0.7322,
674
+ "step": 470
675
+ },
676
+ {
677
+ "epoch": 0.68,
678
+ "grad_norm": 0.6160006018394228,
679
+ "learning_rate": 5.7638085359147235e-06,
680
+ "loss": 0.7235,
681
+ "step": 475
682
+ },
683
+ {
684
+ "epoch": 0.68,
685
+ "grad_norm": 0.6035240580848604,
686
+ "learning_rate": 5.539999136157977e-06,
687
+ "loss": 0.7094,
688
+ "step": 480
689
+ },
690
+ {
691
+ "epoch": 0.69,
692
+ "grad_norm": 0.5166781636185539,
693
+ "learning_rate": 5.318944714707861e-06,
694
+ "loss": 0.7209,
695
+ "step": 485
696
+ },
697
+ {
698
+ "epoch": 0.7,
699
+ "grad_norm": 0.5822772862847312,
700
+ "learning_rate": 5.100781818653549e-06,
701
+ "loss": 0.7088,
702
+ "step": 490
703
+ },
704
+ {
705
+ "epoch": 0.7,
706
+ "grad_norm": 0.6227541043433473,
707
+ "learning_rate": 4.885645208965779e-06,
708
+ "loss": 0.7295,
709
+ "step": 495
710
+ },
711
+ {
712
+ "epoch": 0.71,
713
+ "grad_norm": 0.54846199474894,
714
+ "learning_rate": 4.673667777253944e-06,
715
+ "loss": 0.7452,
716
+ "step": 500
717
+ },
718
+ {
719
+ "epoch": 0.72,
720
+ "grad_norm": 0.6040144782037625,
721
+ "learning_rate": 4.464980463677846e-06,
722
+ "loss": 0.736,
723
+ "step": 505
724
+ },
725
+ {
726
+ "epoch": 0.72,
727
+ "grad_norm": 0.5847697359147894,
728
+ "learning_rate": 4.25971217606493e-06,
729
+ "loss": 0.7364,
730
+ "step": 510
731
+ },
732
+ {
733
+ "epoch": 0.73,
734
+ "grad_norm": 0.597219775666177,
735
+ "learning_rate": 4.057989710282897e-06,
736
+ "loss": 0.7288,
737
+ "step": 515
738
+ },
739
+ {
740
+ "epoch": 0.74,
741
+ "grad_norm": 0.5379233895836751,
742
+ "learning_rate": 3.859937671916833e-06,
743
+ "loss": 0.7383,
744
+ "step": 520
745
+ },
746
+ {
747
+ "epoch": 0.75,
748
+ "grad_norm": 0.6251650029425307,
749
+ "learning_rate": 3.6656783992993885e-06,
750
+ "loss": 0.7264,
751
+ "step": 525
752
+ },
753
+ {
754
+ "epoch": 0.75,
755
+ "grad_norm": 0.6054608357116987,
756
+ "learning_rate": 3.475331887941388e-06,
757
+ "loss": 0.7384,
758
+ "step": 530
759
+ },
760
+ {
761
+ "epoch": 0.76,
762
+ "grad_norm": 0.5551297177270929,
763
+ "learning_rate": 3.2890157164096315e-06,
764
+ "loss": 0.7398,
765
+ "step": 535
766
+ },
767
+ {
768
+ "epoch": 0.77,
769
+ "grad_norm": 0.6396725383662202,
770
+ "learning_rate": 3.1068449736977015e-06,
771
+ "loss": 0.7341,
772
+ "step": 540
773
+ },
774
+ {
775
+ "epoch": 0.77,
776
+ "grad_norm": 0.5648152571609389,
777
+ "learning_rate": 2.9289321881345257e-06,
778
+ "loss": 0.7244,
779
+ "step": 545
780
+ },
781
+ {
782
+ "epoch": 0.78,
783
+ "grad_norm": 0.57252018708692,
784
+ "learning_rate": 2.755387257874764e-06,
785
+ "loss": 0.7228,
786
+ "step": 550
787
+ },
788
+ {
789
+ "epoch": 0.79,
790
+ "grad_norm": 0.5878135364575673,
791
+ "learning_rate": 2.5863173830138212e-06,
792
+ "loss": 0.7181,
793
+ "step": 555
794
+ },
795
+ {
796
+ "epoch": 0.8,
797
+ "grad_norm": 0.5840059710891027,
798
+ "learning_rate": 2.4218269993694733e-06,
799
+ "loss": 0.7286,
800
+ "step": 560
801
+ },
802
+ {
803
+ "epoch": 0.8,
804
+ "grad_norm": 0.5438192269829342,
805
+ "learning_rate": 2.262017713971063e-06,
806
+ "loss": 0.712,
807
+ "step": 565
808
+ },
809
+ {
810
+ "epoch": 0.81,
811
+ "grad_norm": 0.5350040038892513,
812
+ "learning_rate": 2.106988242295981e-06,
813
+ "loss": 0.7312,
814
+ "step": 570
815
+ },
816
+ {
817
+ "epoch": 0.82,
818
+ "grad_norm": 0.5423604435174225,
819
+ "learning_rate": 1.9568343472923524e-06,
820
+ "loss": 0.7155,
821
+ "step": 575
822
+ },
823
+ {
824
+ "epoch": 0.82,
825
+ "grad_norm": 0.6113034501153347,
826
+ "learning_rate": 1.8116487802254868e-06,
827
+ "loss": 0.7282,
828
+ "step": 580
829
+ },
830
+ {
831
+ "epoch": 0.83,
832
+ "grad_norm": 0.6023686669705883,
833
+ "learning_rate": 1.6715212233846656e-06,
834
+ "loss": 0.7159,
835
+ "step": 585
836
+ },
837
+ {
838
+ "epoch": 0.84,
839
+ "grad_norm": 0.5618755550032292,
840
+ "learning_rate": 1.5365382346857005e-06,
841
+ "loss": 0.7349,
842
+ "step": 590
843
+ },
844
+ {
845
+ "epoch": 0.85,
846
+ "grad_norm": 0.5579514856251387,
847
+ "learning_rate": 1.4067831942033904e-06,
848
+ "loss": 0.7106,
849
+ "step": 595
850
+ },
851
+ {
852
+ "epoch": 0.85,
853
+ "grad_norm": 0.5910581925001689,
854
+ "learning_rate": 1.2823362526669825e-06,
855
+ "loss": 0.7074,
856
+ "step": 600
857
+ },
858
+ {
859
+ "epoch": 0.86,
860
+ "grad_norm": 0.5595582019084373,
861
+ "learning_rate": 1.1632742819504406e-06,
862
+ "loss": 0.7244,
863
+ "step": 605
864
+ },
865
+ {
866
+ "epoch": 0.87,
867
+ "grad_norm": 0.6034146234727371,
868
+ "learning_rate": 1.0496708275880497e-06,
869
+ "loss": 0.7237,
870
+ "step": 610
871
+ },
872
+ {
873
+ "epoch": 0.87,
874
+ "grad_norm": 0.5913679291364223,
875
+ "learning_rate": 9.415960633447674e-07,
876
+ "loss": 0.7158,
877
+ "step": 615
878
+ },
879
+ {
880
+ "epoch": 0.88,
881
+ "grad_norm": 0.5694936405039339,
882
+ "learning_rate": 8.391167478693241e-07,
883
+ "loss": 0.7209,
884
+ "step": 620
885
+ },
886
+ {
887
+ "epoch": 0.89,
888
+ "grad_norm": 0.5805765466263503,
889
+ "learning_rate": 7.422961834568565e-07,
890
+ "loss": 0.7097,
891
+ "step": 625
892
+ },
893
+ {
894
+ "epoch": 0.9,
895
+ "grad_norm": 0.5641375110299046,
896
+ "learning_rate": 6.51194176946588e-07,
897
+ "loss": 0.711,
898
+ "step": 630
899
+ },
900
+ {
901
+ "epoch": 0.9,
902
+ "grad_norm": 0.5928177799223828,
903
+ "learning_rate": 5.658670027786561e-07,
904
+ "loss": 0.713,
905
+ "step": 635
906
+ },
907
+ {
908
+ "epoch": 0.91,
909
+ "grad_norm": 0.6022673346018554,
910
+ "learning_rate": 4.863673682329373e-07,
911
+ "loss": 0.7395,
912
+ "step": 640
913
+ },
914
+ {
915
+ "epoch": 0.92,
916
+ "grad_norm": 0.588777849626458,
917
+ "learning_rate": 4.1274438087135273e-07,
918
+ "loss": 0.7435,
919
+ "step": 645
920
+ },
921
+ {
922
+ "epoch": 0.92,
923
+ "grad_norm": 0.52867680060555,
924
+ "learning_rate": 3.450435182037104e-07,
925
+ "loss": 0.6871,
926
+ "step": 650
927
+ },
928
+ {
929
+ "epoch": 0.93,
930
+ "grad_norm": 0.5676798921047626,
931
+ "learning_rate": 2.8330659959589944e-07,
932
+ "loss": 0.727,
933
+ "step": 655
934
+ },
935
+ {
936
+ "epoch": 0.94,
937
+ "grad_norm": 0.5666677388076765,
938
+ "learning_rate": 2.275717604377292e-07,
939
+ "loss": 0.7329,
940
+ "step": 660
941
+ },
942
+ {
943
+ "epoch": 0.95,
944
+ "grad_norm": 0.5389907093110754,
945
+ "learning_rate": 1.7787342858638589e-07,
946
+ "loss": 0.7136,
947
+ "step": 665
948
+ },
949
+ {
950
+ "epoch": 0.95,
951
+ "grad_norm": 0.571817712811333,
952
+ "learning_rate": 1.3424230310007946e-07,
953
+ "loss": 0.739,
954
+ "step": 670
955
+ },
956
+ {
957
+ "epoch": 0.96,
958
+ "grad_norm": 0.5830917030629162,
959
+ "learning_rate": 9.670533527498139e-08,
960
+ "loss": 0.7209,
961
+ "step": 675
962
+ },
963
+ {
964
+ "epoch": 0.97,
965
+ "grad_norm": 0.5837803782222483,
966
+ "learning_rate": 6.528571199719502e-08,
967
+ "loss": 0.7322,
968
+ "step": 680
969
+ },
970
+ {
971
+ "epoch": 0.97,
972
+ "grad_norm": 0.5711509584660898,
973
+ "learning_rate": 4.000284142003264e-08,
974
+ "loss": 0.7113,
975
+ "step": 685
976
+ },
977
+ {
978
+ "epoch": 0.98,
979
+ "grad_norm": 0.5826105062040772,
980
+ "learning_rate": 2.0872340975438555e-08,
981
+ "loss": 0.7327,
982
+ "step": 690
983
+ },
984
+ {
985
+ "epoch": 0.99,
986
+ "grad_norm": 0.5724533629251033,
987
+ "learning_rate": 7.906027726981568e-09,
988
+ "loss": 0.715,
989
+ "step": 695
990
+ },
991
+ {
992
+ "epoch": 1.0,
993
+ "grad_norm": 0.5578776974422313,
994
+ "learning_rate": 1.111911070356131e-09,
995
+ "loss": 0.7072,
996
+ "step": 700
997
+ },
998
+ {
999
+ "epoch": 1.0,
1000
+ "eval_loss": 0.7502214908599854,
1001
+ "eval_runtime": 9.4753,
1002
+ "eval_samples_per_second": 52.769,
1003
+ "eval_steps_per_second": 1.689,
1004
+ "step": 703
1005
+ },
1006
+ {
1007
+ "epoch": 1.0,
1008
+ "step": 703,
1009
+ "total_flos": 101830592102400.0,
1010
+ "train_loss": 0.7627222812701425,
1011
+ "train_runtime": 5862.0034,
1012
+ "train_samples_per_second": 15.353,
1013
+ "train_steps_per_second": 0.12
1014
+ }
1015
+ ],
1016
+ "logging_steps": 5,
1017
+ "max_steps": 703,
1018
+ "num_input_tokens_seen": 0,
1019
+ "num_train_epochs": 1,
1020
+ "save_steps": 100,
1021
+ "total_flos": 101830592102400.0,
1022
+ "train_batch_size": 16,
1023
+ "trial_name": null,
1024
+ "trial_params": null
1025
+ }