kevinoli commited on
Commit
4ff7d03
1 Parent(s): a0469b9

Training in progress, step 11500, checkpoint

Browse files
checkpoint-11500/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1fb30a598d7941b12e4c1cb2798519257260a8f0cba3d94fcd622fe8b2645465
3
  size 1711848436
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c0c7f6e38142e6f510ea1c8befa86f98a0f24681fb50011a329ee87476af7cd1
3
  size 1711848436
checkpoint-11500/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:98335f5e67aa7094e376e0e194c26a65fe256c9bf744c69153b15a04e6550785
3
  size 3424043887
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa2e942c9b03ff7d7752926f89c982bbcb5bb9e551d7e300ea072c868943c79f
3
  size 3424043887
checkpoint-11500/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:338f9b481bd024cab57b03f3e735d52dfcfed2accc305a469a073165e432d332
3
  size 14503
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:27dc0f7dad0524048ee8c15d9f18ec3bc66aeedce9a18091c8b58ac01d452c5c
3
  size 14503
checkpoint-11500/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0d39370aac378b22fae640e92b5c7b87fc2a1f6eca8cc70430d6113b788c8fa7
3
  size 623
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef6d0ef35924c34b39cd5d7515c58d158d26a22d09d10f78ce6920c35a1a9878
3
  size 623
checkpoint-11500/trainer_state.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
- "best_metric": 0.5341849327087402,
3
- "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l57-l/checkpoint-11000",
4
- "epoch": 2.1186440677966103,
5
  "eval_steps": 500,
6
  "global_step": 11500,
7
  "is_hyper_param_search": false,
@@ -9,353 +9,353 @@
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 0.09211495946941783,
13
- "grad_norm": 3.0293657779693604,
14
- "learning_rate": 4.884856300663227e-07,
15
- "loss": 0.271,
16
  "step": 500
17
  },
18
  {
19
- "epoch": 0.09211495946941783,
20
- "eval_loss": 1.069305419921875,
21
- "eval_runtime": 75.1409,
22
- "eval_samples_per_second": 16.063,
23
- "eval_steps_per_second": 2.01,
24
  "step": 500
25
  },
26
  {
27
- "epoch": 0.18422991893883567,
28
- "grad_norm": 7.3040852546691895,
29
- "learning_rate": 4.769712601326456e-07,
30
- "loss": 0.2493,
31
  "step": 1000
32
  },
33
  {
34
- "epoch": 0.18422991893883567,
35
- "eval_loss": 0.9427077770233154,
36
- "eval_runtime": 74.8926,
37
- "eval_samples_per_second": 16.116,
38
- "eval_steps_per_second": 2.016,
39
  "step": 1000
40
  },
41
  {
42
- "epoch": 0.2763448784082535,
43
- "grad_norm": 0.020736657083034515,
44
- "learning_rate": 4.654568901989683e-07,
45
- "loss": 0.2348,
46
  "step": 1500
47
  },
48
  {
49
- "epoch": 0.2763448784082535,
50
- "eval_loss": 0.8726764917373657,
51
- "eval_runtime": 76.6621,
52
- "eval_samples_per_second": 15.744,
53
- "eval_steps_per_second": 1.97,
54
  "step": 1500
55
  },
56
  {
57
- "epoch": 0.36845983787767134,
58
- "grad_norm": 5.3738909628009424e-05,
59
- "learning_rate": 4.5394252026529107e-07,
60
- "loss": 0.1552,
61
  "step": 2000
62
  },
63
  {
64
- "epoch": 0.36845983787767134,
65
- "eval_loss": 0.8325753808021545,
66
- "eval_runtime": 76.6127,
67
- "eval_samples_per_second": 15.755,
68
- "eval_steps_per_second": 1.971,
69
  "step": 2000
70
  },
71
  {
72
- "epoch": 0.46057479734708917,
73
- "grad_norm": 404.510009765625,
74
- "learning_rate": 4.4242815033161386e-07,
75
- "loss": 0.1753,
76
  "step": 2500
77
  },
78
  {
79
- "epoch": 0.46057479734708917,
80
- "eval_loss": 0.7550467848777771,
81
- "eval_runtime": 76.8024,
82
- "eval_samples_per_second": 15.716,
83
- "eval_steps_per_second": 1.966,
84
  "step": 2500
85
  },
86
  {
87
- "epoch": 0.552689756816507,
88
- "grad_norm": 370.7579650878906,
89
- "learning_rate": 4.3091378039793665e-07,
90
- "loss": 0.1659,
91
  "step": 3000
92
  },
93
  {
94
- "epoch": 0.552689756816507,
95
- "eval_loss": 0.719167172908783,
96
- "eval_runtime": 76.6327,
97
- "eval_samples_per_second": 15.75,
98
- "eval_steps_per_second": 1.97,
99
  "step": 3000
100
  },
101
  {
102
- "epoch": 0.6448047162859248,
103
- "grad_norm": 24.97766876220703,
104
- "learning_rate": 4.193994104642594e-07,
105
- "loss": 0.105,
106
  "step": 3500
107
  },
108
  {
109
- "epoch": 0.6448047162859248,
110
- "eval_loss": 0.7117515802383423,
111
- "eval_runtime": 77.1286,
112
- "eval_samples_per_second": 15.649,
113
- "eval_steps_per_second": 1.958,
114
  "step": 3500
115
  },
116
  {
117
- "epoch": 0.7369196757553427,
118
- "grad_norm": 305.62060546875,
119
- "learning_rate": 4.0788504053058217e-07,
120
- "loss": 0.1336,
121
  "step": 4000
122
  },
123
  {
124
- "epoch": 0.7369196757553427,
125
- "eval_loss": 0.6952972412109375,
126
- "eval_runtime": 76.7253,
127
- "eval_samples_per_second": 15.731,
128
- "eval_steps_per_second": 1.968,
129
  "step": 4000
130
  },
131
  {
132
- "epoch": 0.8290346352247605,
133
- "grad_norm": 506.12237548828125,
134
- "learning_rate": 3.9637067059690496e-07,
135
- "loss": 0.1154,
136
  "step": 4500
137
  },
138
  {
139
- "epoch": 0.8290346352247605,
140
- "eval_loss": 0.6744509935379028,
141
- "eval_runtime": 76.9493,
142
- "eval_samples_per_second": 15.686,
143
- "eval_steps_per_second": 1.962,
144
  "step": 4500
145
  },
146
  {
147
- "epoch": 0.9211495946941783,
148
- "grad_norm": 8.823301322991028e-05,
149
- "learning_rate": 3.8485630066322774e-07,
150
- "loss": 0.108,
151
  "step": 5000
152
  },
153
  {
154
- "epoch": 0.9211495946941783,
155
- "eval_loss": 0.6559990048408508,
156
- "eval_runtime": 75.2391,
157
- "eval_samples_per_second": 16.042,
158
- "eval_steps_per_second": 2.007,
159
  "step": 5000
160
  },
161
  {
162
- "epoch": 1.013264554163596,
163
- "grad_norm": 0.0008560372516512871,
164
- "learning_rate": 3.733419307295504e-07,
165
- "loss": 0.1106,
166
  "step": 5500
167
  },
168
  {
169
- "epoch": 1.013264554163596,
170
- "eval_loss": 0.6367496848106384,
171
- "eval_runtime": 77.0031,
172
- "eval_samples_per_second": 15.675,
173
- "eval_steps_per_second": 1.961,
174
  "step": 5500
175
  },
176
  {
177
- "epoch": 1.105379513633014,
178
- "grad_norm": 0.04510360211133957,
179
- "learning_rate": 3.618275607958732e-07,
180
- "loss": 0.0591,
181
  "step": 6000
182
  },
183
  {
184
- "epoch": 1.105379513633014,
185
- "eval_loss": 0.6259325742721558,
186
- "eval_runtime": 75.2787,
187
- "eval_samples_per_second": 16.034,
188
- "eval_steps_per_second": 2.006,
189
  "step": 6000
190
  },
191
  {
192
- "epoch": 1.1974944731024317,
193
- "grad_norm": 0.0005158882704563439,
194
- "learning_rate": 3.50313190862196e-07,
195
- "loss": 0.0745,
196
  "step": 6500
197
  },
198
  {
199
- "epoch": 1.1974944731024317,
200
- "eval_loss": 0.6209793090820312,
201
- "eval_runtime": 77.1989,
202
- "eval_samples_per_second": 15.635,
203
- "eval_steps_per_second": 1.956,
204
  "step": 6500
205
  },
206
  {
207
- "epoch": 1.2896094325718497,
208
- "grad_norm": 3.5883346072296263e-07,
209
- "learning_rate": 3.387988209285188e-07,
210
- "loss": 0.0502,
211
  "step": 7000
212
  },
213
  {
214
- "epoch": 1.2896094325718497,
215
- "eval_loss": 0.6133300065994263,
216
- "eval_runtime": 75.1116,
217
- "eval_samples_per_second": 16.069,
218
- "eval_steps_per_second": 2.01,
219
  "step": 7000
220
  },
221
  {
222
- "epoch": 1.3817243920412676,
223
- "grad_norm": 21.66597557067871,
224
- "learning_rate": 3.272844509948415e-07,
225
- "loss": 0.079,
226
  "step": 7500
227
  },
228
  {
229
- "epoch": 1.3817243920412676,
230
- "eval_loss": 0.6006803512573242,
231
- "eval_runtime": 75.0086,
232
- "eval_samples_per_second": 16.091,
233
- "eval_steps_per_second": 2.013,
234
  "step": 7500
235
  },
236
  {
237
- "epoch": 1.4738393515106853,
238
- "grad_norm": 7.992870814632624e-05,
239
- "learning_rate": 3.157700810611643e-07,
240
- "loss": 0.0776,
241
  "step": 8000
242
  },
243
  {
244
- "epoch": 1.4738393515106853,
245
- "eval_loss": 0.5866427421569824,
246
- "eval_runtime": 75.6702,
247
- "eval_samples_per_second": 15.951,
248
- "eval_steps_per_second": 1.996,
249
  "step": 8000
250
  },
251
  {
252
- "epoch": 1.565954310980103,
253
- "grad_norm": 0.0010381735628470778,
254
- "learning_rate": 3.042557111274871e-07,
255
- "loss": 0.0492,
256
  "step": 8500
257
  },
258
  {
259
- "epoch": 1.565954310980103,
260
- "eval_loss": 0.5679212212562561,
261
- "eval_runtime": 75.0643,
262
- "eval_samples_per_second": 16.08,
263
- "eval_steps_per_second": 2.012,
264
  "step": 8500
265
  },
266
  {
267
- "epoch": 1.658069270449521,
268
- "grad_norm": 1.733377956725235e-07,
269
- "learning_rate": 2.927413411938099e-07,
270
- "loss": 0.0794,
271
  "step": 9000
272
  },
273
  {
274
- "epoch": 1.658069270449521,
275
- "eval_loss": 0.5761673450469971,
276
- "eval_runtime": 76.2738,
277
- "eval_samples_per_second": 15.825,
278
  "eval_steps_per_second": 1.98,
279
  "step": 9000
280
  },
281
  {
282
- "epoch": 1.750184229918939,
283
- "grad_norm": 3.1063287359023306e-11,
284
- "learning_rate": 2.812269712601326e-07,
285
- "loss": 0.0677,
286
  "step": 9500
287
  },
288
  {
289
- "epoch": 1.750184229918939,
290
- "eval_loss": 0.5565799474716187,
291
- "eval_runtime": 75.6897,
292
- "eval_samples_per_second": 15.947,
293
- "eval_steps_per_second": 1.995,
294
  "step": 9500
295
  },
296
  {
297
- "epoch": 1.8422991893883567,
298
- "grad_norm": 0.00010032750287791714,
299
- "learning_rate": 2.697126013264554e-07,
300
- "loss": 0.0566,
301
  "step": 10000
302
  },
303
  {
304
- "epoch": 1.8422991893883567,
305
- "eval_loss": 0.5482199192047119,
306
- "eval_runtime": 75.0029,
307
- "eval_samples_per_second": 16.093,
308
- "eval_steps_per_second": 2.013,
309
  "step": 10000
310
  },
311
  {
312
- "epoch": 1.9344141488577744,
313
- "grad_norm": 5.553330595375883e-08,
314
- "learning_rate": 2.581982313927782e-07,
315
- "loss": 0.0828,
316
  "step": 10500
317
  },
318
  {
319
- "epoch": 1.9344141488577744,
320
- "eval_loss": 0.5500391125679016,
321
- "eval_runtime": 76.2612,
322
- "eval_samples_per_second": 15.827,
323
- "eval_steps_per_second": 1.98,
324
  "step": 10500
325
  },
326
  {
327
- "epoch": 2.026529108327192,
328
- "grad_norm": 1.5537947319899104e-06,
329
- "learning_rate": 2.466838614591009e-07,
330
- "loss": 0.0573,
331
  "step": 11000
332
  },
333
  {
334
- "epoch": 2.026529108327192,
335
- "eval_loss": 0.5341849327087402,
336
- "eval_runtime": 75.6131,
337
- "eval_samples_per_second": 15.963,
338
- "eval_steps_per_second": 1.997,
339
  "step": 11000
340
  },
341
  {
342
- "epoch": 2.1186440677966103,
343
- "grad_norm": 36.10622024536133,
344
- "learning_rate": 2.3516949152542374e-07,
345
- "loss": 0.0401,
346
  "step": 11500
347
  },
348
  {
349
- "epoch": 2.1186440677966103,
350
- "eval_loss": 0.5351412296295166,
351
- "eval_runtime": 74.8873,
352
- "eval_samples_per_second": 16.118,
353
- "eval_steps_per_second": 2.016,
354
  "step": 11500
355
  }
356
  ],
357
  "logging_steps": 500,
358
- "max_steps": 21712,
359
  "num_input_tokens_seen": 0,
360
  "num_train_epochs": 4,
361
  "save_steps": 500,
@@ -371,7 +371,7 @@
371
  "attributes": {}
372
  }
373
  },
374
- "total_flos": 4137940556192520.0,
375
  "train_batch_size": 2,
376
  "trial_name": null,
377
  "trial_params": null
 
1
  {
2
+ "best_metric": 0.6376460790634155,
3
+ "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l57-l/checkpoint-11500",
4
+ "epoch": 1.294607677586401,
5
  "eval_steps": 500,
6
  "global_step": 11500,
7
  "is_hyper_param_search": false,
 
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 0.056287290329843524,
13
+ "grad_norm": 150.1642303466797,
14
+ "learning_rate": 4.929640887087696e-07,
15
+ "loss": 0.3486,
16
  "step": 500
17
  },
18
  {
19
+ "epoch": 0.056287290329843524,
20
+ "eval_loss": 1.1265727281570435,
21
+ "eval_runtime": 123.7806,
22
+ "eval_samples_per_second": 15.948,
23
+ "eval_steps_per_second": 1.995,
24
  "step": 500
25
  },
26
  {
27
+ "epoch": 0.11257458065968705,
28
+ "grad_norm": 366.6913146972656,
29
+ "learning_rate": 4.859281774175391e-07,
30
+ "loss": 0.2733,
31
  "step": 1000
32
  },
33
  {
34
+ "epoch": 0.11257458065968705,
35
+ "eval_loss": 0.9742079377174377,
36
+ "eval_runtime": 125.8047,
37
+ "eval_samples_per_second": 15.691,
38
+ "eval_steps_per_second": 1.963,
39
  "step": 1000
40
  },
41
  {
42
+ "epoch": 0.16886187098953057,
43
+ "grad_norm": 555.5016479492188,
44
+ "learning_rate": 4.788922661263087e-07,
45
+ "loss": 0.1851,
46
  "step": 1500
47
  },
48
  {
49
+ "epoch": 0.16886187098953057,
50
+ "eval_loss": 0.9162159562110901,
51
+ "eval_runtime": 126.7606,
52
+ "eval_samples_per_second": 15.573,
53
+ "eval_steps_per_second": 1.949,
54
  "step": 1500
55
  },
56
  {
57
+ "epoch": 0.2251491613193741,
58
+ "grad_norm": 0.423260897397995,
59
+ "learning_rate": 4.7185635483507824e-07,
60
+ "loss": 0.1973,
61
  "step": 2000
62
  },
63
  {
64
+ "epoch": 0.2251491613193741,
65
+ "eval_loss": 0.8716973662376404,
66
+ "eval_runtime": 127.6955,
67
+ "eval_samples_per_second": 15.459,
68
+ "eval_steps_per_second": 1.934,
69
  "step": 2000
70
  },
71
  {
72
+ "epoch": 0.2814364516492176,
73
+ "grad_norm": 6.840139389038086,
74
+ "learning_rate": 4.6482044354384774e-07,
75
+ "loss": 0.1881,
76
  "step": 2500
77
  },
78
  {
79
+ "epoch": 0.2814364516492176,
80
+ "eval_loss": 0.830047070980072,
81
+ "eval_runtime": 128.0771,
82
+ "eval_samples_per_second": 15.413,
83
+ "eval_steps_per_second": 1.929,
84
  "step": 2500
85
  },
86
  {
87
+ "epoch": 0.33772374197906113,
88
+ "grad_norm": 729.8377685546875,
89
+ "learning_rate": 4.577845322526173e-07,
90
+ "loss": 0.1695,
91
  "step": 3000
92
  },
93
  {
94
+ "epoch": 0.33772374197906113,
95
+ "eval_loss": 0.8265627026557922,
96
+ "eval_runtime": 128.052,
97
+ "eval_samples_per_second": 15.416,
98
+ "eval_steps_per_second": 1.929,
99
  "step": 3000
100
  },
101
  {
102
+ "epoch": 0.39401103230890466,
103
+ "grad_norm": 7.96999938756926e-06,
104
+ "learning_rate": 4.507486209613869e-07,
105
+ "loss": 0.155,
106
  "step": 3500
107
  },
108
  {
109
+ "epoch": 0.39401103230890466,
110
+ "eval_loss": 0.7721038460731506,
111
+ "eval_runtime": 128.1871,
112
+ "eval_samples_per_second": 15.399,
113
+ "eval_steps_per_second": 1.927,
114
  "step": 3500
115
  },
116
  {
117
+ "epoch": 0.4502983226387482,
118
+ "grad_norm": 0.006555848754942417,
119
+ "learning_rate": 4.4371270967015645e-07,
120
+ "loss": 0.1216,
121
  "step": 4000
122
  },
123
  {
124
+ "epoch": 0.4502983226387482,
125
+ "eval_loss": 0.7490401864051819,
126
+ "eval_runtime": 127.5516,
127
+ "eval_samples_per_second": 15.476,
128
+ "eval_steps_per_second": 1.936,
129
  "step": 4000
130
  },
131
  {
132
+ "epoch": 0.5065856129685917,
133
+ "grad_norm": 0.0004515685432124883,
134
+ "learning_rate": 4.36676798378926e-07,
135
+ "loss": 0.1531,
136
  "step": 4500
137
  },
138
  {
139
+ "epoch": 0.5065856129685917,
140
+ "eval_loss": 0.7533182501792908,
141
+ "eval_runtime": 127.4797,
142
+ "eval_samples_per_second": 15.485,
143
+ "eval_steps_per_second": 1.938,
144
  "step": 4500
145
  },
146
  {
147
+ "epoch": 0.5628729032984352,
148
+ "grad_norm": 4.4633176003117114e-05,
149
+ "learning_rate": 4.2964088708769556e-07,
150
+ "loss": 0.1154,
151
  "step": 5000
152
  },
153
  {
154
+ "epoch": 0.5628729032984352,
155
+ "eval_loss": 0.747530460357666,
156
+ "eval_runtime": 127.8675,
157
+ "eval_samples_per_second": 15.438,
158
+ "eval_steps_per_second": 1.932,
159
  "step": 5000
160
  },
161
  {
162
+ "epoch": 0.6191601936282788,
163
+ "grad_norm": 4.158839702606201,
164
+ "learning_rate": 4.2260497579646517e-07,
165
+ "loss": 0.1407,
166
  "step": 5500
167
  },
168
  {
169
+ "epoch": 0.6191601936282788,
170
+ "eval_loss": 0.724854588508606,
171
+ "eval_runtime": 125.1629,
172
+ "eval_samples_per_second": 15.771,
173
+ "eval_steps_per_second": 1.973,
174
  "step": 5500
175
  },
176
  {
177
+ "epoch": 0.6754474839581223,
178
+ "grad_norm": 0.748630166053772,
179
+ "learning_rate": 4.155690645052347e-07,
180
+ "loss": 0.1447,
181
  "step": 6000
182
  },
183
  {
184
+ "epoch": 0.6754474839581223,
185
+ "eval_loss": 0.7091771960258484,
186
+ "eval_runtime": 125.2549,
187
+ "eval_samples_per_second": 15.76,
188
+ "eval_steps_per_second": 1.972,
189
  "step": 6000
190
  },
191
  {
192
+ "epoch": 0.7317347742879657,
193
+ "grad_norm": 2.8120924980612472e-05,
194
+ "learning_rate": 4.0853315321400427e-07,
195
+ "loss": 0.1146,
196
  "step": 6500
197
  },
198
  {
199
+ "epoch": 0.7317347742879657,
200
+ "eval_loss": 0.7105826735496521,
201
+ "eval_runtime": 127.7786,
202
+ "eval_samples_per_second": 15.449,
203
+ "eval_steps_per_second": 1.933,
204
  "step": 6500
205
  },
206
  {
207
+ "epoch": 0.7880220646178093,
208
+ "grad_norm": 1.4741635823156685e-07,
209
+ "learning_rate": 4.014972419227738e-07,
210
+ "loss": 0.171,
211
  "step": 7000
212
  },
213
  {
214
+ "epoch": 0.7880220646178093,
215
+ "eval_loss": 0.6978534460067749,
216
+ "eval_runtime": 125.2567,
217
+ "eval_samples_per_second": 15.76,
218
+ "eval_steps_per_second": 1.972,
219
  "step": 7000
220
  },
221
  {
222
+ "epoch": 0.8443093549476528,
223
+ "grad_norm": 0.002975167240947485,
224
+ "learning_rate": 3.944613306315434e-07,
225
+ "loss": 0.1584,
226
  "step": 7500
227
  },
228
  {
229
+ "epoch": 0.8443093549476528,
230
+ "eval_loss": 0.6923494935035706,
231
+ "eval_runtime": 125.3384,
232
+ "eval_samples_per_second": 15.749,
233
+ "eval_steps_per_second": 1.971,
234
  "step": 7500
235
  },
236
  {
237
+ "epoch": 0.9005966452774964,
238
+ "grad_norm": 9.225498797604814e-05,
239
+ "learning_rate": 3.8742541934031293e-07,
240
+ "loss": 0.1384,
241
  "step": 8000
242
  },
243
  {
244
+ "epoch": 0.9005966452774964,
245
+ "eval_loss": 0.6872902512550354,
246
+ "eval_runtime": 128.114,
247
+ "eval_samples_per_second": 15.408,
248
+ "eval_steps_per_second": 1.928,
249
  "step": 8000
250
  },
251
  {
252
+ "epoch": 0.9568839356073399,
253
+ "grad_norm": 1.9474915902151224e-08,
254
+ "learning_rate": 3.803895080490825e-07,
255
+ "loss": 0.1243,
256
  "step": 8500
257
  },
258
  {
259
+ "epoch": 0.9568839356073399,
260
+ "eval_loss": 0.6788680553436279,
261
+ "eval_runtime": 125.0525,
262
+ "eval_samples_per_second": 15.785,
263
+ "eval_steps_per_second": 1.975,
264
  "step": 8500
265
  },
266
  {
267
+ "epoch": 1.0131712259371835,
268
+ "grad_norm": 5.165647506713867,
269
+ "learning_rate": 3.7335359675785204e-07,
270
+ "loss": 0.1102,
271
  "step": 9000
272
  },
273
  {
274
+ "epoch": 1.0131712259371835,
275
+ "eval_loss": 0.6788864135742188,
276
+ "eval_runtime": 124.7777,
277
+ "eval_samples_per_second": 15.82,
278
  "eval_steps_per_second": 1.98,
279
  "step": 9000
280
  },
281
  {
282
+ "epoch": 1.069458516267027,
283
+ "grad_norm": 0.0005840375670231879,
284
+ "learning_rate": 3.6631768546662164e-07,
285
+ "loss": 0.0523,
286
  "step": 9500
287
  },
288
  {
289
+ "epoch": 1.069458516267027,
290
+ "eval_loss": 0.6628187894821167,
291
+ "eval_runtime": 127.8018,
292
+ "eval_samples_per_second": 15.446,
293
+ "eval_steps_per_second": 1.933,
294
  "step": 9500
295
  },
296
  {
297
+ "epoch": 1.1257458065968704,
298
+ "grad_norm": 0.001191094284877181,
299
+ "learning_rate": 3.592817741753912e-07,
300
+ "loss": 0.0617,
301
  "step": 10000
302
  },
303
  {
304
+ "epoch": 1.1257458065968704,
305
+ "eval_loss": 0.6507942080497742,
306
+ "eval_runtime": 125.484,
307
+ "eval_samples_per_second": 15.731,
308
+ "eval_steps_per_second": 1.968,
309
  "step": 10000
310
  },
311
  {
312
+ "epoch": 1.1820330969267139,
313
+ "grad_norm": 0.0003092484548687935,
314
+ "learning_rate": 3.5224586288416075e-07,
315
+ "loss": 0.0598,
316
  "step": 10500
317
  },
318
  {
319
+ "epoch": 1.1820330969267139,
320
+ "eval_loss": 0.6480408906936646,
321
+ "eval_runtime": 125.7661,
322
+ "eval_samples_per_second": 15.696,
323
+ "eval_steps_per_second": 1.964,
324
  "step": 10500
325
  },
326
  {
327
+ "epoch": 1.2383203872565574,
328
+ "grad_norm": 5.508670710696606e-07,
329
+ "learning_rate": 3.452099515929303e-07,
330
+ "loss": 0.0782,
331
  "step": 11000
332
  },
333
  {
334
+ "epoch": 1.2383203872565574,
335
+ "eval_loss": 0.6403237581253052,
336
+ "eval_runtime": 128.3947,
337
+ "eval_samples_per_second": 15.374,
338
+ "eval_steps_per_second": 1.924,
339
  "step": 11000
340
  },
341
  {
342
+ "epoch": 1.294607677586401,
343
+ "grad_norm": 0.1470593810081482,
344
+ "learning_rate": 3.3817404030169986e-07,
345
+ "loss": 0.0616,
346
  "step": 11500
347
  },
348
  {
349
+ "epoch": 1.294607677586401,
350
+ "eval_loss": 0.6376460790634155,
351
+ "eval_runtime": 125.5391,
352
+ "eval_samples_per_second": 15.724,
353
+ "eval_steps_per_second": 1.968,
354
  "step": 11500
355
  }
356
  ],
357
  "logging_steps": 500,
358
+ "max_steps": 35532,
359
  "num_input_tokens_seen": 0,
360
  "num_train_epochs": 4,
361
  "save_steps": 500,
 
371
  "attributes": {}
372
  }
373
  },
374
+ "total_flos": 4137760677016890.0,
375
  "train_batch_size": 2,
376
  "trial_name": null,
377
  "trial_params": null
checkpoint-11500/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:bf76416e4a857628175385db7f617e54451cd7ffb62c1f9590ee9f30268f8c5b
3
  size 4847
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:319be2a769ef7117b6826908976f080f9082f159f4f0ad75da1a151ad29d9bff
3
  size 4847