bluuluu commited on
Commit
0d33b18
·
verified ·
1 Parent(s): 42e1cfa

Model save

Browse files
README.md ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: Qwen/Qwen2-VL-2B-Instruct
3
+ library_name: transformers
4
+ model_name: Qwen2-VL-2B-Instruct-SFT
5
+ tags:
6
+ - generated_from_trainer
7
+ - trl
8
+ - sft
9
+ licence: license
10
+ ---
11
+
12
+ # Model Card for Qwen2-VL-2B-Instruct-SFT
13
+
14
+ This model is a fine-tuned version of [Qwen/Qwen2-VL-2B-Instruct](https://huggingface.co/Qwen/Qwen2-VL-2B-Instruct).
15
+ It has been trained using [TRL](https://github.com/huggingface/trl).
16
+
17
+ ## Quick start
18
+
19
+ ```python
20
+ from transformers import pipeline
21
+
22
+ question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
23
+ generator = pipeline("text-generation", model="bluuluu/Qwen2-VL-2B-Instruct-SFT", device="cuda")
24
+ output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
25
+ print(output["generated_text"])
26
+ ```
27
+
28
+ ## Training procedure
29
+
30
+ [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/2741919970-hustvl/huggingface/runs/hxbtcd5t)
31
+
32
+
33
+ This model was trained with SFT.
34
+
35
+ ### Framework versions
36
+
37
+ - TRL: 0.14.0
38
+ - Transformers: 4.51.1
39
+ - Pytorch: 2.5.1
40
+ - Datasets: 3.2.0
41
+ - Tokenizers: 0.21.1
42
+
43
+ ## Citations
44
+
45
+
46
+
47
+ Cite TRL as:
48
+
49
+ ```bibtex
50
+ @misc{vonwerra2022trl,
51
+ title = {{TRL: Transformer Reinforcement Learning}},
52
+ author = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallouédec},
53
+ year = 2020,
54
+ journal = {GitHub repository},
55
+ publisher = {GitHub},
56
+ howpublished = {\url{https://github.com/huggingface/trl}}
57
+ }
58
+ ```
all_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 0.9990182603573532,
3
+ "total_flos": 3801667199303680.0,
4
+ "train_loss": 0.13866351419509207,
5
+ "train_runtime": 19174.2073,
6
+ "train_samples": 40742,
7
+ "train_samples_per_second": 2.125,
8
+ "train_steps_per_second": 0.033
9
+ }
chat_template.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "chat_template": "{% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n{% endif %}<|im_start|>{{ message['role'] }}\n{% if message['content'] is string %}{{ message['content'] }}<|im_end|>\n{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>\n{% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant\n{% endif %}"
3
+ }
generation_config.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "attn_implementation": null,
3
+ "bos_token_id": 151643,
4
+ "do_sample": true,
5
+ "eos_token_id": [
6
+ 151645,
7
+ 151643
8
+ ],
9
+ "pad_token_id": 151643,
10
+ "temperature": 0.01,
11
+ "top_k": 1,
12
+ "top_p": 0.001,
13
+ "transformers_version": "4.51.1",
14
+ "use_cache": false
15
+ }
preprocessor_config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_convert_rgb": true,
3
+ "do_normalize": true,
4
+ "do_rescale": true,
5
+ "do_resize": true,
6
+ "image_mean": [
7
+ 0.48145466,
8
+ 0.4578275,
9
+ 0.40821073
10
+ ],
11
+ "image_processor_type": "Qwen2VLImageProcessor",
12
+ "image_std": [
13
+ 0.26862954,
14
+ 0.26130258,
15
+ 0.27577711
16
+ ],
17
+ "max_pixels": 12845056,
18
+ "merge_size": 2,
19
+ "min_pixels": 3136,
20
+ "patch_size": 14,
21
+ "processor_class": "Qwen2VLProcessor",
22
+ "resample": 3,
23
+ "rescale_factor": 0.00392156862745098,
24
+ "size": {
25
+ "longest_edge": 12845056,
26
+ "shortest_edge": 3136
27
+ },
28
+ "temporal_patch_size": 2
29
+ }
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 0.9990182603573532,
3
+ "total_flos": 3801667199303680.0,
4
+ "train_loss": 0.13866351419509207,
5
+ "train_runtime": 19174.2073,
6
+ "train_samples": 40742,
7
+ "train_samples_per_second": 2.125,
8
+ "train_steps_per_second": 0.033
9
+ }
trainer_state.json ADDED
@@ -0,0 +1,940 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.9990182603573532,
6
+ "eval_steps": 500,
7
+ "global_step": 636,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.007853917141174161,
14
+ "grad_norm": 9.454776697912866,
15
+ "learning_rate": 1.25e-06,
16
+ "loss": 1.9975,
17
+ "step": 5
18
+ },
19
+ {
20
+ "epoch": 0.015707834282348322,
21
+ "grad_norm": 7.36591132940049,
22
+ "learning_rate": 2.8125e-06,
23
+ "loss": 1.8813,
24
+ "step": 10
25
+ },
26
+ {
27
+ "epoch": 0.023561751423522483,
28
+ "grad_norm": 7.154615745335707,
29
+ "learning_rate": 4.3750000000000005e-06,
30
+ "loss": 1.4274,
31
+ "step": 15
32
+ },
33
+ {
34
+ "epoch": 0.031415668564696644,
35
+ "grad_norm": 8.822500376414714,
36
+ "learning_rate": 5.9375e-06,
37
+ "loss": 0.8007,
38
+ "step": 20
39
+ },
40
+ {
41
+ "epoch": 0.0392695857058708,
42
+ "grad_norm": 1.5626818502455462,
43
+ "learning_rate": 7.500000000000001e-06,
44
+ "loss": 0.2708,
45
+ "step": 25
46
+ },
47
+ {
48
+ "epoch": 0.04712350284704497,
49
+ "grad_norm": 0.95138015574769,
50
+ "learning_rate": 9.0625e-06,
51
+ "loss": 0.1895,
52
+ "step": 30
53
+ },
54
+ {
55
+ "epoch": 0.054977419988219124,
56
+ "grad_norm": 0.632172201703669,
57
+ "learning_rate": 1.0625e-05,
58
+ "loss": 0.161,
59
+ "step": 35
60
+ },
61
+ {
62
+ "epoch": 0.06283133712939329,
63
+ "grad_norm": 0.6087184175934243,
64
+ "learning_rate": 1.2187500000000001e-05,
65
+ "loss": 0.1514,
66
+ "step": 40
67
+ },
68
+ {
69
+ "epoch": 0.07068525427056745,
70
+ "grad_norm": 0.43002905751035403,
71
+ "learning_rate": 1.375e-05,
72
+ "loss": 0.1405,
73
+ "step": 45
74
+ },
75
+ {
76
+ "epoch": 0.0785391714117416,
77
+ "grad_norm": 0.4316303477126074,
78
+ "learning_rate": 1.5312500000000003e-05,
79
+ "loss": 0.1323,
80
+ "step": 50
81
+ },
82
+ {
83
+ "epoch": 0.08639308855291576,
84
+ "grad_norm": 0.41295406365002396,
85
+ "learning_rate": 1.6875e-05,
86
+ "loss": 0.1232,
87
+ "step": 55
88
+ },
89
+ {
90
+ "epoch": 0.09424700569408993,
91
+ "grad_norm": 0.3869106906435148,
92
+ "learning_rate": 1.84375e-05,
93
+ "loss": 0.1185,
94
+ "step": 60
95
+ },
96
+ {
97
+ "epoch": 0.10210092283526409,
98
+ "grad_norm": 0.38396751864248035,
99
+ "learning_rate": 2e-05,
100
+ "loss": 0.1127,
101
+ "step": 65
102
+ },
103
+ {
104
+ "epoch": 0.10995483997643825,
105
+ "grad_norm": 0.31600318922334614,
106
+ "learning_rate": 1.9996229574120564e-05,
107
+ "loss": 0.1141,
108
+ "step": 70
109
+ },
110
+ {
111
+ "epoch": 0.1178087571176124,
112
+ "grad_norm": 0.2807626851547414,
113
+ "learning_rate": 1.998492113970451e-05,
114
+ "loss": 0.1091,
115
+ "step": 75
116
+ },
117
+ {
118
+ "epoch": 0.12566267425878658,
119
+ "grad_norm": 0.2712214079154975,
120
+ "learning_rate": 1.99660832242746e-05,
121
+ "loss": 0.1062,
122
+ "step": 80
123
+ },
124
+ {
125
+ "epoch": 0.13351659139996072,
126
+ "grad_norm": 0.3514322765970332,
127
+ "learning_rate": 1.99397300332236e-05,
128
+ "loss": 0.1043,
129
+ "step": 85
130
+ },
131
+ {
132
+ "epoch": 0.1413705085411349,
133
+ "grad_norm": 0.2565181736697417,
134
+ "learning_rate": 1.9905881439102222e-05,
135
+ "loss": 0.1012,
136
+ "step": 90
137
+ },
138
+ {
139
+ "epoch": 0.14922442568230906,
140
+ "grad_norm": 0.22648554493196996,
141
+ "learning_rate": 1.9864562966633517e-05,
142
+ "loss": 0.1091,
143
+ "step": 95
144
+ },
145
+ {
146
+ "epoch": 0.1570783428234832,
147
+ "grad_norm": 0.3160797286348635,
148
+ "learning_rate": 1.9815805773465064e-05,
149
+ "loss": 0.1003,
150
+ "step": 100
151
+ },
152
+ {
153
+ "epoch": 0.16493225996465738,
154
+ "grad_norm": 0.24530695236987562,
155
+ "learning_rate": 1.9759646626673445e-05,
156
+ "loss": 0.1054,
157
+ "step": 105
158
+ },
159
+ {
160
+ "epoch": 0.17278617710583152,
161
+ "grad_norm": 0.3308392923974687,
162
+ "learning_rate": 1.9696127875038753e-05,
163
+ "loss": 0.1021,
164
+ "step": 110
165
+ },
166
+ {
167
+ "epoch": 0.1806400942470057,
168
+ "grad_norm": 0.281121787116649,
169
+ "learning_rate": 1.9625297417109982e-05,
170
+ "loss": 0.0967,
171
+ "step": 115
172
+ },
173
+ {
174
+ "epoch": 0.18849401138817987,
175
+ "grad_norm": 0.20020373984472567,
176
+ "learning_rate": 1.954720866508546e-05,
177
+ "loss": 0.1,
178
+ "step": 120
179
+ },
180
+ {
181
+ "epoch": 0.196347928529354,
182
+ "grad_norm": 0.27280707245432356,
183
+ "learning_rate": 1.946192050453549e-05,
184
+ "loss": 0.0999,
185
+ "step": 125
186
+ },
187
+ {
188
+ "epoch": 0.20420184567052818,
189
+ "grad_norm": 0.24726594932319515,
190
+ "learning_rate": 1.936949724999762e-05,
191
+ "loss": 0.0947,
192
+ "step": 130
193
+ },
194
+ {
195
+ "epoch": 0.21205576281170233,
196
+ "grad_norm": 0.2383332440997248,
197
+ "learning_rate": 1.9270008596478008e-05,
198
+ "loss": 0.0982,
199
+ "step": 135
200
+ },
201
+ {
202
+ "epoch": 0.2199096799528765,
203
+ "grad_norm": 0.2203997646125469,
204
+ "learning_rate": 1.916352956689544e-05,
205
+ "loss": 0.1017,
206
+ "step": 140
207
+ },
208
+ {
209
+ "epoch": 0.22776359709405067,
210
+ "grad_norm": 0.2362083987436209,
211
+ "learning_rate": 1.905014045550767e-05,
212
+ "loss": 0.0987,
213
+ "step": 145
214
+ },
215
+ {
216
+ "epoch": 0.2356175142352248,
217
+ "grad_norm": 0.23431067680898818,
218
+ "learning_rate": 1.89299267673627e-05,
219
+ "loss": 0.1055,
220
+ "step": 150
221
+ },
222
+ {
223
+ "epoch": 0.24347143137639898,
224
+ "grad_norm": 0.21183859562424995,
225
+ "learning_rate": 1.88029791538207e-05,
226
+ "loss": 0.0972,
227
+ "step": 155
228
+ },
229
+ {
230
+ "epoch": 0.25132534851757315,
231
+ "grad_norm": 0.22397328492306617,
232
+ "learning_rate": 1.8669393344195154e-05,
233
+ "loss": 0.1009,
234
+ "step": 160
235
+ },
236
+ {
237
+ "epoch": 0.2591792656587473,
238
+ "grad_norm": 0.22197633586551052,
239
+ "learning_rate": 1.852927007356481e-05,
240
+ "loss": 0.0936,
241
+ "step": 165
242
+ },
243
+ {
244
+ "epoch": 0.26703318279992144,
245
+ "grad_norm": 0.21347022910458285,
246
+ "learning_rate": 1.8382715006810853e-05,
247
+ "loss": 0.0957,
248
+ "step": 170
249
+ },
250
+ {
251
+ "epoch": 0.2748870999410956,
252
+ "grad_norm": 0.19957938768669062,
253
+ "learning_rate": 1.8229838658936566e-05,
254
+ "loss": 0.0873,
255
+ "step": 175
256
+ },
257
+ {
258
+ "epoch": 0.2827410170822698,
259
+ "grad_norm": 0.21372555100198343,
260
+ "learning_rate": 1.807075631172963e-05,
261
+ "loss": 0.0931,
262
+ "step": 180
263
+ },
264
+ {
265
+ "epoch": 0.29059493422344396,
266
+ "grad_norm": 0.24190910676139596,
267
+ "learning_rate": 1.7905587926829815e-05,
268
+ "loss": 0.0954,
269
+ "step": 185
270
+ },
271
+ {
272
+ "epoch": 0.29844885136461813,
273
+ "grad_norm": 0.2348573879122653,
274
+ "learning_rate": 1.77344580552677e-05,
275
+ "loss": 0.092,
276
+ "step": 190
277
+ },
278
+ {
279
+ "epoch": 0.30630276850579224,
280
+ "grad_norm": 0.21204667498616597,
281
+ "learning_rate": 1.7557495743542586e-05,
282
+ "loss": 0.0865,
283
+ "step": 195
284
+ },
285
+ {
286
+ "epoch": 0.3141566856469664,
287
+ "grad_norm": 0.23025633594185083,
288
+ "learning_rate": 1.7374834436310427e-05,
289
+ "loss": 0.0892,
290
+ "step": 200
291
+ },
292
+ {
293
+ "epoch": 0.3220106027881406,
294
+ "grad_norm": 0.1958893222829977,
295
+ "learning_rate": 1.7186611875755227e-05,
296
+ "loss": 0.0896,
297
+ "step": 205
298
+ },
299
+ {
300
+ "epoch": 0.32986451992931476,
301
+ "grad_norm": 0.22589475386438357,
302
+ "learning_rate": 1.6992969997719658e-05,
303
+ "loss": 0.0945,
304
+ "step": 210
305
+ },
306
+ {
307
+ "epoch": 0.33771843707048893,
308
+ "grad_norm": 0.20789776750177996,
309
+ "learning_rate": 1.679405482467338e-05,
310
+ "loss": 0.0904,
311
+ "step": 215
312
+ },
313
+ {
314
+ "epoch": 0.34557235421166305,
315
+ "grad_norm": 0.22053512123077823,
316
+ "learning_rate": 1.6590016355599653e-05,
317
+ "loss": 0.0882,
318
+ "step": 220
319
+ },
320
+ {
321
+ "epoch": 0.3534262713528372,
322
+ "grad_norm": 0.20437871058955417,
323
+ "learning_rate": 1.638100845288331e-05,
324
+ "loss": 0.0943,
325
+ "step": 225
326
+ },
327
+ {
328
+ "epoch": 0.3612801884940114,
329
+ "grad_norm": 0.2090640789388261,
330
+ "learning_rate": 1.6167188726285433e-05,
331
+ "loss": 0.0924,
332
+ "step": 230
333
+ },
334
+ {
335
+ "epoch": 0.36913410563518556,
336
+ "grad_norm": 0.21993457381708645,
337
+ "learning_rate": 1.5948718414092163e-05,
338
+ "loss": 0.082,
339
+ "step": 235
340
+ },
341
+ {
342
+ "epoch": 0.37698802277635973,
343
+ "grad_norm": 0.21951962440906275,
344
+ "learning_rate": 1.5725762261527295e-05,
345
+ "loss": 0.0918,
346
+ "step": 240
347
+ },
348
+ {
349
+ "epoch": 0.38484193991753385,
350
+ "grad_norm": 0.19872240145820674,
351
+ "learning_rate": 1.549848839652035e-05,
352
+ "loss": 0.0955,
353
+ "step": 245
354
+ },
355
+ {
356
+ "epoch": 0.392695857058708,
357
+ "grad_norm": 0.2184861301177773,
358
+ "learning_rate": 1.5267068202923802e-05,
359
+ "loss": 0.0899,
360
+ "step": 250
361
+ },
362
+ {
363
+ "epoch": 0.4005497741998822,
364
+ "grad_norm": 0.21059689355945896,
365
+ "learning_rate": 1.503167619127504e-05,
366
+ "loss": 0.0919,
367
+ "step": 255
368
+ },
369
+ {
370
+ "epoch": 0.40840369134105636,
371
+ "grad_norm": 0.21063064163571898,
372
+ "learning_rate": 1.479248986720057e-05,
373
+ "loss": 0.086,
374
+ "step": 260
375
+ },
376
+ {
377
+ "epoch": 0.41625760848223053,
378
+ "grad_norm": 0.19969666720995202,
379
+ "learning_rate": 1.4549689597561652e-05,
380
+ "loss": 0.0949,
381
+ "step": 265
382
+ },
383
+ {
384
+ "epoch": 0.42411152562340465,
385
+ "grad_norm": 0.21154526388191275,
386
+ "learning_rate": 1.4303458474442325e-05,
387
+ "loss": 0.0869,
388
+ "step": 270
389
+ },
390
+ {
391
+ "epoch": 0.4319654427645788,
392
+ "grad_norm": 0.2305997701221425,
393
+ "learning_rate": 1.4053982177082369e-05,
394
+ "loss": 0.09,
395
+ "step": 275
396
+ },
397
+ {
398
+ "epoch": 0.439819359905753,
399
+ "grad_norm": 0.18766215420046584,
400
+ "learning_rate": 1.3801448831859363e-05,
401
+ "loss": 0.0869,
402
+ "step": 280
403
+ },
404
+ {
405
+ "epoch": 0.44767327704692716,
406
+ "grad_norm": 0.22942333663793876,
407
+ "learning_rate": 1.3546048870425356e-05,
408
+ "loss": 0.0929,
409
+ "step": 285
410
+ },
411
+ {
412
+ "epoch": 0.45552719418810134,
413
+ "grad_norm": 0.24300158612250203,
414
+ "learning_rate": 1.328797488610519e-05,
415
+ "loss": 0.0865,
416
+ "step": 290
417
+ },
418
+ {
419
+ "epoch": 0.46338111132927545,
420
+ "grad_norm": 0.20752136799187096,
421
+ "learning_rate": 1.3027421488664723e-05,
422
+ "loss": 0.0927,
423
+ "step": 295
424
+ },
425
+ {
426
+ "epoch": 0.4712350284704496,
427
+ "grad_norm": 0.21217812864504576,
428
+ "learning_rate": 1.2764585157558486e-05,
429
+ "loss": 0.0879,
430
+ "step": 300
431
+ },
432
+ {
433
+ "epoch": 0.4790889456116238,
434
+ "grad_norm": 0.22299949763483096,
435
+ "learning_rate": 1.2499664093767458e-05,
436
+ "loss": 0.0938,
437
+ "step": 305
438
+ },
439
+ {
440
+ "epoch": 0.48694286275279797,
441
+ "grad_norm": 0.23957557508813998,
442
+ "learning_rate": 1.2232858070338618e-05,
443
+ "loss": 0.0904,
444
+ "step": 310
445
+ },
446
+ {
447
+ "epoch": 0.49479677989397214,
448
+ "grad_norm": 0.2023486493539765,
449
+ "learning_rate": 1.1964368281739078e-05,
450
+ "loss": 0.0874,
451
+ "step": 315
452
+ },
453
+ {
454
+ "epoch": 0.5026506970351463,
455
+ "grad_norm": 0.1861038104854032,
456
+ "learning_rate": 1.1694397192138295e-05,
457
+ "loss": 0.0816,
458
+ "step": 320
459
+ },
460
+ {
461
+ "epoch": 0.5105046141763204,
462
+ "grad_norm": 0.2014117982786214,
463
+ "learning_rate": 1.1423148382732854e-05,
464
+ "loss": 0.0823,
465
+ "step": 325
466
+ },
467
+ {
468
+ "epoch": 0.5183585313174947,
469
+ "grad_norm": 0.22960664067039807,
470
+ "learning_rate": 1.1150826398228904e-05,
471
+ "loss": 0.0882,
472
+ "step": 330
473
+ },
474
+ {
475
+ "epoch": 0.5262124484586688,
476
+ "grad_norm": 0.222023007250355,
477
+ "learning_rate": 1.087763659259803e-05,
478
+ "loss": 0.0863,
479
+ "step": 335
480
+ },
481
+ {
482
+ "epoch": 0.5340663655998429,
483
+ "grad_norm": 0.19468935638754573,
484
+ "learning_rate": 1.0603784974222862e-05,
485
+ "loss": 0.0884,
486
+ "step": 340
487
+ },
488
+ {
489
+ "epoch": 0.5419202827410171,
490
+ "grad_norm": 0.2047369290635767,
491
+ "learning_rate": 1.0329478050549208e-05,
492
+ "loss": 0.0781,
493
+ "step": 345
494
+ },
495
+ {
496
+ "epoch": 0.5497741998821912,
497
+ "grad_norm": 0.2057117950843339,
498
+ "learning_rate": 1.0054922672361858e-05,
499
+ "loss": 0.0813,
500
+ "step": 350
501
+ },
502
+ {
503
+ "epoch": 0.5576281170233655,
504
+ "grad_norm": 0.23197127290798342,
505
+ "learning_rate": 9.780325877801455e-06,
506
+ "loss": 0.0989,
507
+ "step": 355
508
+ },
509
+ {
510
+ "epoch": 0.5654820341645396,
511
+ "grad_norm": 0.3305907734524743,
512
+ "learning_rate": 9.50589473624013e-06,
513
+ "loss": 0.0816,
514
+ "step": 360
515
+ },
516
+ {
517
+ "epoch": 0.5733359513057137,
518
+ "grad_norm": 0.215453345630245,
519
+ "learning_rate": 9.231836192133532e-06,
520
+ "loss": 0.0861,
521
+ "step": 365
522
+ },
523
+ {
524
+ "epoch": 0.5811898684468879,
525
+ "grad_norm": 0.1908748794995548,
526
+ "learning_rate": 8.958356908967104e-06,
527
+ "loss": 0.0855,
528
+ "step": 370
529
+ },
530
+ {
531
+ "epoch": 0.589043785588062,
532
+ "grad_norm": 0.28572080285977014,
533
+ "learning_rate": 8.685663113414186e-06,
534
+ "loss": 0.0841,
535
+ "step": 375
536
+ },
537
+ {
538
+ "epoch": 0.5968977027292363,
539
+ "grad_norm": 0.29366381772227595,
540
+ "learning_rate": 8.413960439823567e-06,
541
+ "loss": 0.0824,
542
+ "step": 380
543
+ },
544
+ {
545
+ "epoch": 0.6047516198704104,
546
+ "grad_norm": 0.21387609796609025,
547
+ "learning_rate": 8.143453775153646e-06,
548
+ "loss": 0.0846,
549
+ "step": 385
550
+ },
551
+ {
552
+ "epoch": 0.6126055370115845,
553
+ "grad_norm": 0.18382031478587926,
554
+ "learning_rate": 7.874347104470234e-06,
555
+ "loss": 0.0886,
556
+ "step": 390
557
+ },
558
+ {
559
+ "epoch": 0.6204594541527587,
560
+ "grad_norm": 0.20443500947333,
561
+ "learning_rate": 7.606843357124426e-06,
562
+ "loss": 0.0818,
563
+ "step": 395
564
+ },
565
+ {
566
+ "epoch": 0.6283133712939328,
567
+ "grad_norm": 0.1868753348119755,
568
+ "learning_rate": 7.341144253726583e-06,
569
+ "loss": 0.0801,
570
+ "step": 400
571
+ },
572
+ {
573
+ "epoch": 0.6361672884351071,
574
+ "grad_norm": 0.1919110684349949,
575
+ "learning_rate": 7.0774501540318305e-06,
576
+ "loss": 0.0837,
577
+ "step": 405
578
+ },
579
+ {
580
+ "epoch": 0.6440212055762812,
581
+ "grad_norm": 0.19733646385316747,
582
+ "learning_rate": 6.815959905851715e-06,
583
+ "loss": 0.0907,
584
+ "step": 410
585
+ },
586
+ {
587
+ "epoch": 0.6518751227174553,
588
+ "grad_norm": 0.20695118469945245,
589
+ "learning_rate": 6.556870695106028e-06,
590
+ "loss": 0.0839,
591
+ "step": 415
592
+ },
593
+ {
594
+ "epoch": 0.6597290398586295,
595
+ "grad_norm": 0.206342467457721,
596
+ "learning_rate": 6.300377897127825e-06,
597
+ "loss": 0.0817,
598
+ "step": 420
599
+ },
600
+ {
601
+ "epoch": 0.6675829569998036,
602
+ "grad_norm": 0.19498340314833404,
603
+ "learning_rate": 6.046674929333787e-06,
604
+ "loss": 0.0832,
605
+ "step": 425
606
+ },
607
+ {
608
+ "epoch": 0.6754368741409779,
609
+ "grad_norm": 0.18423257358883185,
610
+ "learning_rate": 5.795953105371e-06,
611
+ "loss": 0.0894,
612
+ "step": 430
613
+ },
614
+ {
615
+ "epoch": 0.683290791282152,
616
+ "grad_norm": 0.19500438735567713,
617
+ "learning_rate": 5.548401490850193e-06,
618
+ "loss": 0.0841,
619
+ "step": 435
620
+ },
621
+ {
622
+ "epoch": 0.6911447084233261,
623
+ "grad_norm": 0.1943344540910271,
624
+ "learning_rate": 5.304206760774139e-06,
625
+ "loss": 0.0857,
626
+ "step": 440
627
+ },
628
+ {
629
+ "epoch": 0.6989986255645003,
630
+ "grad_norm": 0.18694442634566882,
631
+ "learning_rate": 5.063553058768814e-06,
632
+ "loss": 0.0873,
633
+ "step": 445
634
+ },
635
+ {
636
+ "epoch": 0.7068525427056744,
637
+ "grad_norm": 0.2528592095445329,
638
+ "learning_rate": 4.826621858223431e-06,
639
+ "loss": 0.0817,
640
+ "step": 450
641
+ },
642
+ {
643
+ "epoch": 0.7147064598468487,
644
+ "grad_norm": 0.2090763236109062,
645
+ "learning_rate": 4.593591825444028e-06,
646
+ "loss": 0.0845,
647
+ "step": 455
648
+ },
649
+ {
650
+ "epoch": 0.7225603769880228,
651
+ "grad_norm": 0.18944053970106092,
652
+ "learning_rate": 4.364638684923848e-06,
653
+ "loss": 0.0871,
654
+ "step": 460
655
+ },
656
+ {
657
+ "epoch": 0.7304142941291969,
658
+ "grad_norm": 0.20174647482444266,
659
+ "learning_rate": 4.13993508683214e-06,
660
+ "loss": 0.084,
661
+ "step": 465
662
+ },
663
+ {
664
+ "epoch": 0.7382682112703711,
665
+ "grad_norm": 0.19890794882086518,
666
+ "learning_rate": 3.919650476821192e-06,
667
+ "loss": 0.0848,
668
+ "step": 470
669
+ },
670
+ {
671
+ "epoch": 0.7461221284115452,
672
+ "grad_norm": 0.17866134537019918,
673
+ "learning_rate": 3.7039509682498887e-06,
674
+ "loss": 0.0767,
675
+ "step": 475
676
+ },
677
+ {
678
+ "epoch": 0.7539760455527195,
679
+ "grad_norm": 0.1758570225271436,
680
+ "learning_rate": 3.4929992169200865e-06,
681
+ "loss": 0.079,
682
+ "step": 480
683
+ },
684
+ {
685
+ "epoch": 0.7618299626938936,
686
+ "grad_norm": 0.19532675849601197,
687
+ "learning_rate": 3.2869542984202974e-06,
688
+ "loss": 0.0872,
689
+ "step": 485
690
+ },
691
+ {
692
+ "epoch": 0.7696838798350677,
693
+ "grad_norm": 0.18336677624616388,
694
+ "learning_rate": 3.0859715881691267e-06,
695
+ "loss": 0.078,
696
+ "step": 490
697
+ },
698
+ {
699
+ "epoch": 0.7775377969762419,
700
+ "grad_norm": 0.19477530024566625,
701
+ "learning_rate": 2.890202644248983e-06,
702
+ "loss": 0.0888,
703
+ "step": 495
704
+ },
705
+ {
706
+ "epoch": 0.785391714117416,
707
+ "grad_norm": 0.18612540878842743,
708
+ "learning_rate": 2.6997950931183736e-06,
709
+ "loss": 0.0806,
710
+ "step": 500
711
+ },
712
+ {
713
+ "epoch": 0.785391714117416,
714
+ "eval_loss": 0.07095803320407867,
715
+ "eval_runtime": 199.4165,
716
+ "eval_samples_per_second": 9.046,
717
+ "eval_steps_per_second": 2.262,
718
+ "step": 500
719
+ },
720
+ {
721
+ "epoch": 0.7932456312585903,
722
+ "grad_norm": 0.20861974562159583,
723
+ "learning_rate": 2.514892518288988e-06,
724
+ "loss": 0.0816,
725
+ "step": 505
726
+ },
727
+ {
728
+ "epoch": 0.8010995483997644,
729
+ "grad_norm": 0.19828891809733562,
730
+ "learning_rate": 2.335634352051488e-06,
731
+ "loss": 0.0927,
732
+ "step": 510
733
+ },
734
+ {
735
+ "epoch": 0.8089534655409385,
736
+ "grad_norm": 0.17567422510168001,
737
+ "learning_rate": 2.1621557703316876e-06,
738
+ "loss": 0.077,
739
+ "step": 515
740
+ },
741
+ {
742
+ "epoch": 0.8168073826821127,
743
+ "grad_norm": 0.43901734050274105,
744
+ "learning_rate": 1.994587590756397e-06,
745
+ "loss": 0.0879,
746
+ "step": 520
747
+ },
748
+ {
749
+ "epoch": 0.8246612998232868,
750
+ "grad_norm": 0.19303609197478824,
751
+ "learning_rate": 1.8330561740057839e-06,
752
+ "loss": 0.0798,
753
+ "step": 525
754
+ },
755
+ {
756
+ "epoch": 0.8325152169644611,
757
+ "grad_norm": 0.21014700571002942,
758
+ "learning_rate": 1.6776833285266602e-06,
759
+ "loss": 0.0846,
760
+ "step": 530
761
+ },
762
+ {
763
+ "epoch": 0.8403691341056352,
764
+ "grad_norm": 0.19381351445718112,
765
+ "learning_rate": 1.528586218678535e-06,
766
+ "loss": 0.0828,
767
+ "step": 535
768
+ },
769
+ {
770
+ "epoch": 0.8482230512468093,
771
+ "grad_norm": 0.20006096517086086,
772
+ "learning_rate": 1.3858772763817174e-06,
773
+ "loss": 0.0817,
774
+ "step": 540
775
+ },
776
+ {
777
+ "epoch": 0.8560769683879835,
778
+ "grad_norm": 0.25891905601753346,
779
+ "learning_rate": 1.2496641163340562e-06,
780
+ "loss": 0.0815,
781
+ "step": 545
782
+ },
783
+ {
784
+ "epoch": 0.8639308855291576,
785
+ "grad_norm": 0.20338109286639236,
786
+ "learning_rate": 1.120049454860307e-06,
787
+ "loss": 0.0827,
788
+ "step": 550
789
+ },
790
+ {
791
+ "epoch": 0.8717848026703319,
792
+ "grad_norm": 0.1817547896016754,
793
+ "learning_rate": 9.971310324552597e-07,
794
+ "loss": 0.0845,
795
+ "step": 555
796
+ },
797
+ {
798
+ "epoch": 0.879638719811506,
799
+ "grad_norm": 0.20333004229909032,
800
+ "learning_rate": 8.810015400790994e-07,
801
+ "loss": 0.0836,
802
+ "step": 560
803
+ },
804
+ {
805
+ "epoch": 0.8874926369526801,
806
+ "grad_norm": 0.2005258856819357,
807
+ "learning_rate": 7.71748549260507e-07,
808
+ "loss": 0.0813,
809
+ "step": 565
810
+ },
811
+ {
812
+ "epoch": 0.8953465540938543,
813
+ "grad_norm": 0.17505423789580266,
814
+ "learning_rate": 6.694544460602825e-07,
815
+ "loss": 0.0881,
816
+ "step": 570
817
+ },
818
+ {
819
+ "epoch": 0.9032004712350284,
820
+ "grad_norm": 0.19509475957323322,
821
+ "learning_rate": 5.741963689452268e-07,
822
+ "loss": 0.082,
823
+ "step": 575
824
+ },
825
+ {
826
+ "epoch": 0.9110543883762027,
827
+ "grad_norm": 0.19518986043069664,
828
+ "learning_rate": 4.860461506191782e-07,
829
+ "loss": 0.0842,
830
+ "step": 580
831
+ },
832
+ {
833
+ "epoch": 0.9189083055173768,
834
+ "grad_norm": 0.1925064249529951,
835
+ "learning_rate": 4.0507026385502747e-07,
836
+ "loss": 0.0839,
837
+ "step": 585
838
+ },
839
+ {
840
+ "epoch": 0.9267622226585509,
841
+ "grad_norm": 0.18327816158141888,
842
+ "learning_rate": 3.313297713685859e-07,
843
+ "loss": 0.087,
844
+ "step": 590
845
+ },
846
+ {
847
+ "epoch": 0.9346161397997251,
848
+ "grad_norm": 0.20588287191995575,
849
+ "learning_rate": 2.6488027977210175e-07,
850
+ "loss": 0.0827,
851
+ "step": 595
852
+ },
853
+ {
854
+ "epoch": 0.9424700569408992,
855
+ "grad_norm": 0.197357389547204,
856
+ "learning_rate": 2.057718976421341e-07,
857
+ "loss": 0.0815,
858
+ "step": 600
859
+ },
860
+ {
861
+ "epoch": 0.9503239740820735,
862
+ "grad_norm": 0.18513291000922844,
863
+ "learning_rate": 1.5404919773341576e-07,
864
+ "loss": 0.082,
865
+ "step": 605
866
+ },
867
+ {
868
+ "epoch": 0.9581778912232476,
869
+ "grad_norm": 0.185473668497204,
870
+ "learning_rate": 1.0975118336720603e-07,
871
+ "loss": 0.0846,
872
+ "step": 610
873
+ },
874
+ {
875
+ "epoch": 0.9660318083644217,
876
+ "grad_norm": 0.1803986934893818,
877
+ "learning_rate": 7.291125901946027e-08,
878
+ "loss": 0.0777,
879
+ "step": 615
880
+ },
881
+ {
882
+ "epoch": 0.9738857255055959,
883
+ "grad_norm": 0.18308958490608873,
884
+ "learning_rate": 4.3557205131008475e-08,
885
+ "loss": 0.0771,
886
+ "step": 620
887
+ },
888
+ {
889
+ "epoch": 0.98173964264677,
890
+ "grad_norm": 0.2968041785813041,
891
+ "learning_rate": 2.171115715874139e-08,
892
+ "loss": 0.0872,
893
+ "step": 625
894
+ },
895
+ {
896
+ "epoch": 0.9895935597879443,
897
+ "grad_norm": 0.19838566337698643,
898
+ "learning_rate": 7.389588883585097e-09,
899
+ "loss": 0.0789,
900
+ "step": 630
901
+ },
902
+ {
903
+ "epoch": 0.9974474769291184,
904
+ "grad_norm": 0.17786747719054677,
905
+ "learning_rate": 6.032999878735624e-10,
906
+ "loss": 0.0821,
907
+ "step": 635
908
+ },
909
+ {
910
+ "epoch": 0.9990182603573532,
911
+ "step": 636,
912
+ "total_flos": 3801667199303680.0,
913
+ "train_loss": 0.13866351419509207,
914
+ "train_runtime": 19174.2073,
915
+ "train_samples_per_second": 2.125,
916
+ "train_steps_per_second": 0.033
917
+ }
918
+ ],
919
+ "logging_steps": 5,
920
+ "max_steps": 636,
921
+ "num_input_tokens_seen": 0,
922
+ "num_train_epochs": 1,
923
+ "save_steps": 1000,
924
+ "stateful_callbacks": {
925
+ "TrainerControl": {
926
+ "args": {
927
+ "should_epoch_stop": false,
928
+ "should_evaluate": false,
929
+ "should_log": false,
930
+ "should_save": true,
931
+ "should_training_stop": true
932
+ },
933
+ "attributes": {}
934
+ }
935
+ },
936
+ "total_flos": 3801667199303680.0,
937
+ "train_batch_size": 1,
938
+ "trial_name": null,
939
+ "trial_params": null
940
+ }