Hugging Face's logo

Spaces:
hbXNov
/
owl-con-demo
Runtime error

App Files Community
1
Hritik commited on
Commit
0187095
•
1 Parent(s): 0eaade4

add owl-con ckpt

Browse files
Files changed (23) hide show
  1. mplugowl7bvideo/config.json +299 -0
  2. mplugowl7bvideo/generation_config.json +11 -0
  3. mplugowl7bvideo/old_pytorch_model.bin +3 -0
  4. mplugowl7bvideo/preprocessor_config.json +9 -0
  5. mplugowl7bvideo/pytorch_model.bin +3 -0
  6. mplugowl7bvideo/tokenizer.model +3 -0
  7. mplugowl7bvideo/tokenizer_config.json +1 -0
  8. owl-con/README.md +6 -0
  9. owl-con/adapter_config.json +21 -0
  10. owl-con/adapter_model.bin +3 -0
  11. owl-con/checkpoint-5178/optimizer.pt +3 -0
  12. owl-con/checkpoint-5178/pytorch_model.bin +3 -0
  13. owl-con/checkpoint-5178/rng_state_0.pth +3 -0
  14. owl-con/checkpoint-5178/rng_state_1.pth +3 -0
  15. owl-con/checkpoint-5178/rng_state_2.pth +3 -0
  16. owl-con/checkpoint-5178/rng_state_3.pth +3 -0
  17. owl-con/checkpoint-5178/scheduler.pt +3 -0
  18. owl-con/checkpoint-5178/special_tokens_map.json +30 -0
  19. owl-con/checkpoint-5178/tokenizer.model +3 -0
  20. owl-con/checkpoint-5178/tokenizer_config.json +40 -0
  21. owl-con/checkpoint-5178/trainer_state.json +3526 -0
  22. owl-con/checkpoint-5178/training_args.bin +3 -0
  23. owl-con/params.txt +36 -0
mplugowl7bvideo/config.json ADDED
@@ -0,0 +1,299 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "return_dict": true,
3
+ "output_hidden_states": false,
4
+ "output_attentions": false,
5
+ "torchscript": false,
6
+ "torch_dtype": null,
7
+ "use_bfloat16": false,
8
+ "tf_legacy_loss": false,
9
+ "pruned_heads": {},
10
+ "tie_word_embeddings": false,
11
+ "is_encoder_decoder": false,
12
+ "is_decoder": false,
13
+ "cross_attention_hidden_size": null,
14
+ "add_cross_attention": false,
15
+ "tie_encoder_decoder": false,
16
+ "max_length": 20,
17
+ "min_length": 0,
18
+ "do_sample": false,
19
+ "early_stopping": false,
20
+ "num_beams": 1,
21
+ "num_beam_groups": 1,
22
+ "diversity_penalty": 0.0,
23
+ "temperature": 1.0,
24
+ "top_k": 50,
25
+ "top_p": 1.0,
26
+ "typical_p": 1.0,
27
+ "repetition_penalty": 1.0,
28
+ "length_penalty": 1.0,
29
+ "no_repeat_ngram_size": 0,
30
+ "encoder_no_repeat_ngram_size": 0,
31
+ "bad_words_ids": null,
32
+ "num_return_sequences": 1,
33
+ "chunk_size_feed_forward": 0,
34
+ "output_scores": false,
35
+ "return_dict_in_generate": false,
36
+ "forced_bos_token_id": null,
37
+ "forced_eos_token_id": null,
38
+ "remove_invalid_values": false,
39
+ "exponential_decay_length_penalty": null,
40
+ "suppress_tokens": null,
41
+ "begin_suppress_tokens": null,
42
+ "architectures": null,
43
+ "finetuning_task": null,
44
+ "id2label": {
45
+ "0": "LABEL_0",
46
+ "1": "LABEL_1"
47
+ },
48
+ "label2id": {
49
+ "LABEL_0": 0,
50
+ "LABEL_1": 1
51
+ },
52
+ "tokenizer_class": null,
53
+ "prefix": null,
54
+ "bos_token_id": null,
55
+ "pad_token_id": null,
56
+ "eos_token_id": null,
57
+ "sep_token_id": null,
58
+ "decoder_start_token_id": null,
59
+ "task_specific_params": null,
60
+ "problem_type": null,
61
+ "_name_or_path": "",
62
+ "_commit_hash": null,
63
+ "transformers_version": null,
64
+ "vision_config": {
65
+ "return_dict": true,
66
+ "output_hidden_states": false,
67
+ "output_attentions": false,
68
+ "torchscript": false,
69
+ "torch_dtype": null,
70
+ "use_bfloat16": false,
71
+ "tf_legacy_loss": false,
72
+ "pruned_heads": {},
73
+ "tie_word_embeddings": true,
74
+ "is_encoder_decoder": false,
75
+ "is_decoder": false,
76
+ "cross_attention_hidden_size": null,
77
+ "add_cross_attention": false,
78
+ "tie_encoder_decoder": false,
79
+ "max_length": 20,
80
+ "min_length": 0,
81
+ "do_sample": false,
82
+ "early_stopping": false,
83
+ "num_beams": 1,
84
+ "num_beam_groups": 1,
85
+ "diversity_penalty": 0.0,
86
+ "temperature": 1.0,
87
+ "top_k": 50,
88
+ "top_p": 1.0,
89
+ "typical_p": 1.0,
90
+ "repetition_penalty": 1.0,
91
+ "length_penalty": 1.0,
92
+ "no_repeat_ngram_size": 0,
93
+ "encoder_no_repeat_ngram_size": 0,
94
+ "bad_words_ids": null,
95
+ "num_return_sequences": 1,
96
+ "chunk_size_feed_forward": 0,
97
+ "output_scores": false,
98
+ "return_dict_in_generate": false,
99
+ "forced_bos_token_id": null,
100
+ "forced_eos_token_id": null,
101
+ "remove_invalid_values": false,
102
+ "exponential_decay_length_penalty": null,
103
+ "suppress_tokens": null,
104
+ "begin_suppress_tokens": null,
105
+ "architectures": null,
106
+ "finetuning_task": null,
107
+ "id2label": {
108
+ "0": "LABEL_0",
109
+ "1": "LABEL_1"
110
+ },
111
+ "label2id": {
112
+ "LABEL_0": 0,
113
+ "LABEL_1": 1
114
+ },
115
+ "tokenizer_class": null,
116
+ "prefix": null,
117
+ "bos_token_id": null,
118
+ "pad_token_id": null,
119
+ "eos_token_id": null,
120
+ "sep_token_id": null,
121
+ "decoder_start_token_id": null,
122
+ "task_specific_params": null,
123
+ "problem_type": null,
124
+ "_name_or_path": "",
125
+ "transformers_version": "4.29.0.dev0",
126
+ "model_type": "mplug_owl_vision_model",
127
+ "hidden_size": 1024,
128
+ "intermediate_size": 4096,
129
+ "projection_dim": 768,
130
+ "num_hidden_layers": 24,
131
+ "num_attention_heads": 16,
132
+ "num_channels": 3,
133
+ "patch_size": 14,
134
+ "image_size": 224,
135
+ "initializer_range": 0.02,
136
+ "initializer_factor": 1.0,
137
+ "attention_dropout": 0.0,
138
+ "layer_norm_eps": 1e-06,
139
+ "hidden_act": "quick_gelu"
140
+ },
141
+ "visual_abstractor_config": {
142
+ "return_dict": true,
143
+ "output_hidden_states": false,
144
+ "output_attentions": false,
145
+ "torchscript": false,
146
+ "torch_dtype": null,
147
+ "use_bfloat16": false,
148
+ "tf_legacy_loss": false,
149
+ "pruned_heads": {},
150
+ "tie_word_embeddings": true,
151
+ "is_encoder_decoder": false,
152
+ "is_decoder": false,
153
+ "cross_attention_hidden_size": null,
154
+ "add_cross_attention": false,
155
+ "tie_encoder_decoder": false,
156
+ "max_length": 20,
157
+ "min_length": 0,
158
+ "do_sample": false,
159
+ "early_stopping": false,
160
+ "num_beams": 1,
161
+ "num_beam_groups": 1,
162
+ "diversity_penalty": 0.0,
163
+ "temperature": 1.0,
164
+ "top_k": 50,
165
+ "top_p": 1.0,
166
+ "typical_p": 1.0,
167
+ "repetition_penalty": 1.0,
168
+ "length_penalty": 1.0,
169
+ "no_repeat_ngram_size": 0,
170
+ "encoder_no_repeat_ngram_size": 0,
171
+ "bad_words_ids": null,
172
+ "num_return_sequences": 1,
173
+ "chunk_size_feed_forward": 0,
174
+ "output_scores": false,
175
+ "return_dict_in_generate": false,
176
+ "forced_bos_token_id": null,
177
+ "forced_eos_token_id": null,
178
+ "remove_invalid_values": false,
179
+ "exponential_decay_length_penalty": null,
180
+ "suppress_tokens": null,
181
+ "begin_suppress_tokens": null,
182
+ "architectures": null,
183
+ "finetuning_task": null,
184
+ "id2label": {
185
+ "0": "LABEL_0",
186
+ "1": "LABEL_1"
187
+ },
188
+ "label2id": {
189
+ "LABEL_0": 0,
190
+ "LABEL_1": 1
191
+ },
192
+ "tokenizer_class": null,
193
+ "prefix": null,
194
+ "bos_token_id": null,
195
+ "pad_token_id": 0,
196
+ "eos_token_id": null,
197
+ "sep_token_id": null,
198
+ "decoder_start_token_id": null,
199
+ "task_specific_params": null,
200
+ "problem_type": null,
201
+ "_name_or_path": "",
202
+ "transformers_version": "4.29.0.dev0",
203
+ "vocab_size": 30522,
204
+ "hidden_size": 1024,
205
+ "num_hidden_layers": 6,
206
+ "num_attention_heads": 16,
207
+ "hidden_act": "gelu",
208
+ "intermediate_size": 4096,
209
+ "hidden_dropout_prob": 0.1,
210
+ "attention_probs_dropout_prob": 0.1,
211
+ "max_position_embeddings": 512,
212
+ "initializer_range": 0.02,
213
+ "layer_norm_eps": 1e-06,
214
+ "position_embedding_type": "absolute",
215
+ "classifier_dropout": null,
216
+ "cross_attention_frequency": 2,
217
+ "encoder_hidden_size": 1024,
218
+ "model_type": "MPlugOwlVisualAbstractor"
219
+ },
220
+ "text_config": {
221
+ "vocab_size": 32000,
222
+ "max_position_embeddings": 2048,
223
+ "hidden_size": 4096,
224
+ "intermediate_size": 11008,
225
+ "num_hidden_layers": 32,
226
+ "num_attention_heads": 32,
227
+ "hidden_act": "silu",
228
+ "initializer_range": 0.02,
229
+ "rms_norm_eps": 1e-06,
230
+ "use_cache": true,
231
+ "return_dict": true,
232
+ "output_hidden_states": false,
233
+ "output_attentions": false,
234
+ "torchscript": false,
235
+ "torch_dtype": null,
236
+ "use_bfloat16": false,
237
+ "tf_legacy_loss": false,
238
+ "pruned_heads": {},
239
+ "tie_word_embeddings": false,
240
+ "is_encoder_decoder": false,
241
+ "is_decoder": false,
242
+ "cross_attention_hidden_size": null,
243
+ "add_cross_attention": false,
244
+ "tie_encoder_decoder": false,
245
+ "max_length": 20,
246
+ "min_length": 0,
247
+ "do_sample": false,
248
+ "early_stopping": false,
249
+ "num_beams": 1,
250
+ "num_beam_groups": 1,
251
+ "diversity_penalty": 0.0,
252
+ "temperature": 1.0,
253
+ "top_k": 50,
254
+ "top_p": 1.0,
255
+ "typical_p": 1.0,
256
+ "repetition_penalty": 1.0,
257
+ "length_penalty": 1.0,
258
+ "no_repeat_ngram_size": 0,
259
+ "encoder_no_repeat_ngram_size": 0,
260
+ "bad_words_ids": null,
261
+ "num_return_sequences": 1,
262
+ "chunk_size_feed_forward": 0,
263
+ "output_scores": false,
264
+ "return_dict_in_generate": false,
265
+ "forced_bos_token_id": null,
266
+ "forced_eos_token_id": null,
267
+ "remove_invalid_values": false,
268
+ "exponential_decay_length_penalty": null,
269
+ "suppress_tokens": null,
270
+ "begin_suppress_tokens": null,
271
+ "architectures": null,
272
+ "finetuning_task": null,
273
+ "id2label": {
274
+ "0": "LABEL_0",
275
+ "1": "LABEL_1"
276
+ },
277
+ "label2id": {
278
+ "LABEL_0": 0,
279
+ "LABEL_1": 1
280
+ },
281
+ "tokenizer_class": null,
282
+ "prefix": null,
283
+ "bos_token_id": 1,
284
+ "pad_token_id": 2,
285
+ "eos_token_id": 2,
286
+ "sep_token_id": null,
287
+ "decoder_start_token_id": null,
288
+ "task_specific_params": null,
289
+ "problem_type": null,
290
+ "_name_or_path": "",
291
+ "transformers_version": "4.29.0.dev0",
292
+ "model_type": "llama"
293
+ },
294
+ "num_query_tokens": 64,
295
+ "use_decoder_only_language_model": true,
296
+ "initializer_factor": 1.0,
297
+ "initializer_range": 0.02,
298
+ "model_type": "mplug-owl"
299
+ }
mplugowl7bvideo/generation_config.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 1,
3
+ "do_sample": true,
4
+ "eos_token_id": 2,
5
+ "max_length": 512,
6
+ "pad_token_id": 0,
7
+ "top_k": 3,
8
+ "transformers_version": "4.29.0.dev0",
9
+ "unk_token_id": 0
10
+ }
11
+
mplugowl7bvideo/old_pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:553029b486c521ceec301881cbec0556873e1fb0191267e40292eb061cbf5d40
3
+ size 14305584703
mplugowl7bvideo/preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "resolution": 224,
3
+ "image_mean": [
4
+ 0.48145466, 0.4578275, 0.40821073
5
+ ],
6
+ "image_std": [
7
+ 0.26862954, 0.26130258, 0.27577711
8
+ ]
9
+ }
mplugowl7bvideo/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7bf9c821075524e0ef7692552814a67de82c112d7341a2e7d6603ec51a13421b
3
+ size 14305584703
mplugowl7bvideo/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
mplugowl7bvideo/tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": "<s>", "eos_token": "</s>", "model_max_length": 1000000000000000019884624838656, "tokenizer_class": "LlamaTokenizer", "unk_token": "<unk>", "pad_token": "<unk>"}
owl-con/README.md ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ ## Training procedure
2
+
3
+ ### Framework versions
4
+
5
+
6
+ - PEFT 0.5.0
owl-con/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": {
3
+ "base_model_class": "MplugOwlForConditionalGeneration",
4
+ "parent_library": "mplug_owl_video.modeling_mplug_owl"
5
+ },
6
+ "base_model_name_or_path": "/local2/hbansal/video_text/mplug-owl-llama-7b-video",
7
+ "bias": "none",
8
+ "fan_in_fan_out": false,
9
+ "inference_mode": true,
10
+ "init_lora_weights": true,
11
+ "layers_pattern": null,
12
+ "layers_to_transform": null,
13
+ "lora_alpha": 32,
14
+ "lora_dropout": 0.05,
15
+ "modules_to_save": null,
16
+ "peft_type": "LORA",
17
+ "r": 32,
18
+ "revision": null,
19
+ "target_modules": ".*language_model.*\\.(q_proj|v_proj|k_proj|o_proj|gate_proj|down_proj|up_proj)",
20
+ "task_type": null
21
+ }
owl-con/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:24418f746189910a0cda8440616b8b30069dd2dd3198bdac96e5f9c73cc7194c
3
+ size 319983949
owl-con/checkpoint-5178/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cf1a86302d333ad1cafaad3eb7cc16919a84cbc3625da0671cf679e9ddcc243e
3
+ size 639897797
owl-con/checkpoint-5178/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1f4072beeee705a35d216a3766e51db7c650e072db92fb18804b37c4fc66ed33
3
+ size 14625892902
owl-con/checkpoint-5178/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3600cac96a5f6090f0388e68ce5154e2499b0925fd9a4f9acbf92e14efd041a7
3
+ size 14583
owl-con/checkpoint-5178/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c93cf10ed6fd3c0e772b9b7dfe2a4c963feec3edfe935f8b26221ed13ce8ce3f
3
+ size 14583
owl-con/checkpoint-5178/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:84de293d1b76db69337a844e9fcbb7c3d02cc31f298f66012e021957e93fab45
3
+ size 14583
owl-con/checkpoint-5178/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dcba752ae8c97fd37f217e2e833bd34841568dd8e736506e4b43e031f3286ffe
3
+ size 14583
owl-con/checkpoint-5178/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4364ab5259a0621078e4e8636f22de7cb70a3dda6e1270d6764c760c58aed389
3
+ size 627
owl-con/checkpoint-5178/special_tokens_map.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "<unk>",
18
+ "lstrip": false,
19
+ "normalized": true,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "unk_token": {
24
+ "content": "<unk>",
25
+ "lstrip": false,
26
+ "normalized": true,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ }
30
+ }
owl-con/checkpoint-5178/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
owl-con/checkpoint-5178/tokenizer_config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "bos_token": {
5
+ "__type": "AddedToken",
6
+ "content": "<s>",
7
+ "lstrip": false,
8
+ "normalized": true,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "clean_up_tokenization_spaces": false,
13
+ "eos_token": {
14
+ "__type": "AddedToken",
15
+ "content": "</s>",
16
+ "lstrip": false,
17
+ "normalized": true,
18
+ "rstrip": false,
19
+ "single_word": false
20
+ },
21
+ "model_max_length": 1000000000000000019884624838656,
22
+ "pad_token": {
23
+ "__type": "AddedToken",
24
+ "content": "<unk>",
25
+ "lstrip": false,
26
+ "normalized": true,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "sp_model_kwargs": {},
31
+ "tokenizer_class": "LlamaTokenizer",
32
+ "unk_token": {
33
+ "__type": "AddedToken",
34
+ "content": "<unk>",
35
+ "lstrip": false,
36
+ "normalized": true,
37
+ "rstrip": false,
38
+ "single_word": false
39
+ }
40
+ }
owl-con/checkpoint-5178/trainer_state.json ADDED
@@ -0,0 +1,3526 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.0,
5
+ "global_step": 5178,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 5e-06,
13
+ "loss": 3.0388,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 1e-05,
19
+ "loss": 2.9144,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 1.5e-05,
25
+ "loss": 2.4056,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.02,
30
+ "learning_rate": 2e-05,
31
+ "loss": 1.3107,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.02,
36
+ "learning_rate": 2.5e-05,
37
+ "loss": 0.8564,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.02,
42
+ "learning_rate": 3e-05,
43
+ "loss": 0.6775,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.03,
48
+ "learning_rate": 3.5e-05,
49
+ "loss": 0.54,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.03,
54
+ "learning_rate": 4e-05,
55
+ "loss": 0.5361,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.03,
60
+ "learning_rate": 4.5e-05,
61
+ "loss": 0.4847,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.04,
66
+ "learning_rate": 5e-05,
67
+ "loss": 0.4707,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.04,
72
+ "eval_loss": 0.4903041422367096,
73
+ "eval_runtime": 241.1648,
74
+ "eval_samples_per_second": 2.073,
75
+ "eval_steps_per_second": 0.066,
76
+ "step": 100
77
+ },
78
+ {
79
+ "epoch": 0.04,
80
+ "learning_rate": 5.500000000000001e-05,
81
+ "loss": 0.4843,
82
+ "step": 110
83
+ },
84
+ {
85
+ "epoch": 0.05,
86
+ "learning_rate": 6e-05,
87
+ "loss": 0.4912,
88
+ "step": 120
89
+ },
90
+ {
91
+ "epoch": 0.05,
92
+ "learning_rate": 6.500000000000001e-05,
93
+ "loss": 0.453,
94
+ "step": 130
95
+ },
96
+ {
97
+ "epoch": 0.05,
98
+ "learning_rate": 7e-05,
99
+ "loss": 0.4396,
100
+ "step": 140
101
+ },
102
+ {
103
+ "epoch": 0.06,
104
+ "learning_rate": 7.500000000000001e-05,
105
+ "loss": 0.4066,
106
+ "step": 150
107
+ },
108
+ {
109
+ "epoch": 0.06,
110
+ "learning_rate": 8e-05,
111
+ "loss": 0.4339,
112
+ "step": 160
113
+ },
114
+ {
115
+ "epoch": 0.07,
116
+ "learning_rate": 8.5e-05,
117
+ "loss": 0.4042,
118
+ "step": 170
119
+ },
120
+ {
121
+ "epoch": 0.07,
122
+ "learning_rate": 9e-05,
123
+ "loss": 0.4036,
124
+ "step": 180
125
+ },
126
+ {
127
+ "epoch": 0.07,
128
+ "learning_rate": 9.5e-05,
129
+ "loss": 0.4017,
130
+ "step": 190
131
+ },
132
+ {
133
+ "epoch": 0.08,
134
+ "learning_rate": 0.0001,
135
+ "loss": 0.4132,
136
+ "step": 200
137
+ },
138
+ {
139
+ "epoch": 0.08,
140
+ "eval_loss": 0.4102957546710968,
141
+ "eval_runtime": 237.7984,
142
+ "eval_samples_per_second": 2.103,
143
+ "eval_steps_per_second": 0.067,
144
+ "step": 200
145
+ },
146
+ {
147
+ "epoch": 0.08,
148
+ "learning_rate": 9.979911611088791e-05,
149
+ "loss": 0.4073,
150
+ "step": 210
151
+ },
152
+ {
153
+ "epoch": 0.08,
154
+ "learning_rate": 9.959823222177582e-05,
155
+ "loss": 0.3744,
156
+ "step": 220
157
+ },
158
+ {
159
+ "epoch": 0.09,
160
+ "learning_rate": 9.939734833266372e-05,
161
+ "loss": 0.4091,
162
+ "step": 230
163
+ },
164
+ {
165
+ "epoch": 0.09,
166
+ "learning_rate": 9.919646444355163e-05,
167
+ "loss": 0.3846,
168
+ "step": 240
169
+ },
170
+ {
171
+ "epoch": 0.1,
172
+ "learning_rate": 9.899558055443953e-05,
173
+ "loss": 0.4056,
174
+ "step": 250
175
+ },
176
+ {
177
+ "epoch": 0.1,
178
+ "learning_rate": 9.879469666532744e-05,
179
+ "loss": 0.3713,
180
+ "step": 260
181
+ },
182
+ {
183
+ "epoch": 0.1,
184
+ "learning_rate": 9.859381277621536e-05,
185
+ "loss": 0.3851,
186
+ "step": 270
187
+ },
188
+ {
189
+ "epoch": 0.11,
190
+ "learning_rate": 9.839292888710326e-05,
191
+ "loss": 0.3802,
192
+ "step": 280
193
+ },
194
+ {
195
+ "epoch": 0.11,
196
+ "learning_rate": 9.819204499799116e-05,
197
+ "loss": 0.3883,
198
+ "step": 290
199
+ },
200
+ {
201
+ "epoch": 0.12,
202
+ "learning_rate": 9.799116110887908e-05,
203
+ "loss": 0.3939,
204
+ "step": 300
205
+ },
206
+ {
207
+ "epoch": 0.12,
208
+ "eval_loss": 0.3881138265132904,
209
+ "eval_runtime": 233.0966,
210
+ "eval_samples_per_second": 2.145,
211
+ "eval_steps_per_second": 0.069,
212
+ "step": 300
213
+ },
214
+ {
215
+ "epoch": 0.12,
216
+ "learning_rate": 9.779027721976698e-05,
217
+ "loss": 0.3892,
218
+ "step": 310
219
+ },
220
+ {
221
+ "epoch": 0.12,
222
+ "learning_rate": 9.758939333065489e-05,
223
+ "loss": 0.3568,
224
+ "step": 320
225
+ },
226
+ {
227
+ "epoch": 0.13,
228
+ "learning_rate": 9.73885094415428e-05,
229
+ "loss": 0.3924,
230
+ "step": 330
231
+ },
232
+ {
233
+ "epoch": 0.13,
234
+ "learning_rate": 9.71876255524307e-05,
235
+ "loss": 0.3665,
236
+ "step": 340
237
+ },
238
+ {
239
+ "epoch": 0.14,
240
+ "learning_rate": 9.69867416633186e-05,
241
+ "loss": 0.3685,
242
+ "step": 350
243
+ },
244
+ {
245
+ "epoch": 0.14,
246
+ "learning_rate": 9.678585777420651e-05,
247
+ "loss": 0.3713,
248
+ "step": 360
249
+ },
250
+ {
251
+ "epoch": 0.14,
252
+ "learning_rate": 9.658497388509442e-05,
253
+ "loss": 0.356,
254
+ "step": 370
255
+ },
256
+ {
257
+ "epoch": 0.15,
258
+ "learning_rate": 9.638408999598232e-05,
259
+ "loss": 0.3484,
260
+ "step": 380
261
+ },
262
+ {
263
+ "epoch": 0.15,
264
+ "learning_rate": 9.618320610687024e-05,
265
+ "loss": 0.3865,
266
+ "step": 390
267
+ },
268
+ {
269
+ "epoch": 0.15,
270
+ "learning_rate": 9.598232221775813e-05,
271
+ "loss": 0.3645,
272
+ "step": 400
273
+ },
274
+ {
275
+ "epoch": 0.15,
276
+ "eval_loss": 0.3724190890789032,
277
+ "eval_runtime": 233.3567,
278
+ "eval_samples_per_second": 2.143,
279
+ "eval_steps_per_second": 0.069,
280
+ "step": 400
281
+ },
282
+ {
283
+ "epoch": 0.16,
284
+ "learning_rate": 9.578143832864604e-05,
285
+ "loss": 0.3665,
286
+ "step": 410
287
+ },
288
+ {
289
+ "epoch": 0.16,
290
+ "learning_rate": 9.558055443953396e-05,
291
+ "loss": 0.3496,
292
+ "step": 420
293
+ },
294
+ {
295
+ "epoch": 0.17,
296
+ "learning_rate": 9.537967055042187e-05,
297
+ "loss": 0.332,
298
+ "step": 430
299
+ },
300
+ {
301
+ "epoch": 0.17,
302
+ "learning_rate": 9.517878666130977e-05,
303
+ "loss": 0.3624,
304
+ "step": 440
305
+ },
306
+ {
307
+ "epoch": 0.17,
308
+ "learning_rate": 9.497790277219766e-05,
309
+ "loss": 0.3503,
310
+ "step": 450
311
+ },
312
+ {
313
+ "epoch": 0.18,
314
+ "learning_rate": 9.477701888308558e-05,
315
+ "loss": 0.337,
316
+ "step": 460
317
+ },
318
+ {
319
+ "epoch": 0.18,
320
+ "learning_rate": 9.457613499397349e-05,
321
+ "loss": 0.376,
322
+ "step": 470
323
+ },
324
+ {
325
+ "epoch": 0.19,
326
+ "learning_rate": 9.43752511048614e-05,
327
+ "loss": 0.3304,
328
+ "step": 480
329
+ },
330
+ {
331
+ "epoch": 0.19,
332
+ "learning_rate": 9.41743672157493e-05,
333
+ "loss": 0.3644,
334
+ "step": 490
335
+ },
336
+ {
337
+ "epoch": 0.19,
338
+ "learning_rate": 9.39734833266372e-05,
339
+ "loss": 0.3618,
340
+ "step": 500
341
+ },
342
+ {
343
+ "epoch": 0.19,
344
+ "eval_loss": 0.3597237765789032,
345
+ "eval_runtime": 233.4866,
346
+ "eval_samples_per_second": 2.141,
347
+ "eval_steps_per_second": 0.069,
348
+ "step": 500
349
+ },
350
+ {
351
+ "epoch": 0.2,
352
+ "learning_rate": 9.377259943752511e-05,
353
+ "loss": 0.3855,
354
+ "step": 510
355
+ },
356
+ {
357
+ "epoch": 0.2,
358
+ "learning_rate": 9.357171554841302e-05,
359
+ "loss": 0.3652,
360
+ "step": 520
361
+ },
362
+ {
363
+ "epoch": 0.2,
364
+ "learning_rate": 9.337083165930092e-05,
365
+ "loss": 0.3621,
366
+ "step": 530
367
+ },
368
+ {
369
+ "epoch": 0.21,
370
+ "learning_rate": 9.316994777018884e-05,
371
+ "loss": 0.3318,
372
+ "step": 540
373
+ },
374
+ {
375
+ "epoch": 0.21,
376
+ "learning_rate": 9.296906388107675e-05,
377
+ "loss": 0.3409,
378
+ "step": 550
379
+ },
380
+ {
381
+ "epoch": 0.22,
382
+ "learning_rate": 9.276817999196466e-05,
383
+ "loss": 0.3336,
384
+ "step": 560
385
+ },
386
+ {
387
+ "epoch": 0.22,
388
+ "learning_rate": 9.256729610285255e-05,
389
+ "loss": 0.3505,
390
+ "step": 570
391
+ },
392
+ {
393
+ "epoch": 0.22,
394
+ "learning_rate": 9.236641221374047e-05,
395
+ "loss": 0.3354,
396
+ "step": 580
397
+ },
398
+ {
399
+ "epoch": 0.23,
400
+ "learning_rate": 9.216552832462837e-05,
401
+ "loss": 0.3619,
402
+ "step": 590
403
+ },
404
+ {
405
+ "epoch": 0.23,
406
+ "learning_rate": 9.196464443551628e-05,
407
+ "loss": 0.3491,
408
+ "step": 600
409
+ },
410
+ {
411
+ "epoch": 0.23,
412
+ "eval_loss": 0.3491559624671936,
413
+ "eval_runtime": 227.9072,
414
+ "eval_samples_per_second": 2.194,
415
+ "eval_steps_per_second": 0.07,
416
+ "step": 600
417
+ },
418
+ {
419
+ "epoch": 0.24,
420
+ "learning_rate": 9.176376054640418e-05,
421
+ "loss": 0.353,
422
+ "step": 610
423
+ },
424
+ {
425
+ "epoch": 0.24,
426
+ "learning_rate": 9.156287665729209e-05,
427
+ "loss": 0.3508,
428
+ "step": 620
429
+ },
430
+ {
431
+ "epoch": 0.24,
432
+ "learning_rate": 9.136199276818e-05,
433
+ "loss": 0.3637,
434
+ "step": 630
435
+ },
436
+ {
437
+ "epoch": 0.25,
438
+ "learning_rate": 9.11611088790679e-05,
439
+ "loss": 0.3297,
440
+ "step": 640
441
+ },
442
+ {
443
+ "epoch": 0.25,
444
+ "learning_rate": 9.096022498995581e-05,
445
+ "loss": 0.3578,
446
+ "step": 650
447
+ },
448
+ {
449
+ "epoch": 0.25,
450
+ "learning_rate": 9.075934110084371e-05,
451
+ "loss": 0.3327,
452
+ "step": 660
453
+ },
454
+ {
455
+ "epoch": 0.26,
456
+ "learning_rate": 9.055845721173163e-05,
457
+ "loss": 0.3336,
458
+ "step": 670
459
+ },
460
+ {
461
+ "epoch": 0.26,
462
+ "learning_rate": 9.035757332261952e-05,
463
+ "loss": 0.3113,
464
+ "step": 680
465
+ },
466
+ {
467
+ "epoch": 0.27,
468
+ "learning_rate": 9.015668943350743e-05,
469
+ "loss": 0.3278,
470
+ "step": 690
471
+ },
472
+ {
473
+ "epoch": 0.27,
474
+ "learning_rate": 8.995580554439535e-05,
475
+ "loss": 0.3358,
476
+ "step": 700
477
+ },
478
+ {
479
+ "epoch": 0.27,
480
+ "eval_loss": 0.3494350016117096,
481
+ "eval_runtime": 226.5818,
482
+ "eval_samples_per_second": 2.207,
483
+ "eval_steps_per_second": 0.071,
484
+ "step": 700
485
+ },
486
+ {
487
+ "epoch": 0.27,
488
+ "learning_rate": 8.975492165528326e-05,
489
+ "loss": 0.3417,
490
+ "step": 710
491
+ },
492
+ {
493
+ "epoch": 0.28,
494
+ "learning_rate": 8.955403776617116e-05,
495
+ "loss": 0.3451,
496
+ "step": 720
497
+ },
498
+ {
499
+ "epoch": 0.28,
500
+ "learning_rate": 8.935315387705905e-05,
501
+ "loss": 0.3298,
502
+ "step": 730
503
+ },
504
+ {
505
+ "epoch": 0.29,
506
+ "learning_rate": 8.915226998794697e-05,
507
+ "loss": 0.3188,
508
+ "step": 740
509
+ },
510
+ {
511
+ "epoch": 0.29,
512
+ "learning_rate": 8.895138609883488e-05,
513
+ "loss": 0.3211,
514
+ "step": 750
515
+ },
516
+ {
517
+ "epoch": 0.29,
518
+ "learning_rate": 8.875050220972279e-05,
519
+ "loss": 0.3324,
520
+ "step": 760
521
+ },
522
+ {
523
+ "epoch": 0.3,
524
+ "learning_rate": 8.854961832061069e-05,
525
+ "loss": 0.3269,
526
+ "step": 770
527
+ },
528
+ {
529
+ "epoch": 0.3,
530
+ "learning_rate": 8.83487344314986e-05,
531
+ "loss": 0.3308,
532
+ "step": 780
533
+ },
534
+ {
535
+ "epoch": 0.31,
536
+ "learning_rate": 8.81478505423865e-05,
537
+ "loss": 0.341,
538
+ "step": 790
539
+ },
540
+ {
541
+ "epoch": 0.31,
542
+ "learning_rate": 8.794696665327441e-05,
543
+ "loss": 0.3299,
544
+ "step": 800
545
+ },
546
+ {
547
+ "epoch": 0.31,
548
+ "eval_loss": 0.3420061469078064,
549
+ "eval_runtime": 222.718,
550
+ "eval_samples_per_second": 2.245,
551
+ "eval_steps_per_second": 0.072,
552
+ "step": 800
553
+ },
554
+ {
555
+ "epoch": 0.31,
556
+ "learning_rate": 8.774608276416231e-05,
557
+ "loss": 0.3289,
558
+ "step": 810
559
+ },
560
+ {
561
+ "epoch": 0.32,
562
+ "learning_rate": 8.754519887505023e-05,
563
+ "loss": 0.328,
564
+ "step": 820
565
+ },
566
+ {
567
+ "epoch": 0.32,
568
+ "learning_rate": 8.734431498593814e-05,
569
+ "loss": 0.3464,
570
+ "step": 830
571
+ },
572
+ {
573
+ "epoch": 0.32,
574
+ "learning_rate": 8.714343109682603e-05,
575
+ "loss": 0.3227,
576
+ "step": 840
577
+ },
578
+ {
579
+ "epoch": 0.33,
580
+ "learning_rate": 8.694254720771394e-05,
581
+ "loss": 0.352,
582
+ "step": 850
583
+ },
584
+ {
585
+ "epoch": 0.33,
586
+ "learning_rate": 8.674166331860186e-05,
587
+ "loss": 0.3555,
588
+ "step": 860
589
+ },
590
+ {
591
+ "epoch": 0.34,
592
+ "learning_rate": 8.654077942948976e-05,
593
+ "loss": 0.3322,
594
+ "step": 870
595
+ },
596
+ {
597
+ "epoch": 0.34,
598
+ "learning_rate": 8.633989554037767e-05,
599
+ "loss": 0.316,
600
+ "step": 880
601
+ },
602
+ {
603
+ "epoch": 0.34,
604
+ "learning_rate": 8.613901165126557e-05,
605
+ "loss": 0.304,
606
+ "step": 890
607
+ },
608
+ {
609
+ "epoch": 0.35,
610
+ "learning_rate": 8.593812776215348e-05,
611
+ "loss": 0.3257,
612
+ "step": 900
613
+ },
614
+ {
615
+ "epoch": 0.35,
616
+ "eval_loss": 0.3412737250328064,
617
+ "eval_runtime": 230.8687,
618
+ "eval_samples_per_second": 2.166,
619
+ "eval_steps_per_second": 0.069,
620
+ "step": 900
621
+ },
622
+ {
623
+ "epoch": 0.35,
624
+ "learning_rate": 8.573724387304139e-05,
625
+ "loss": 0.3339,
626
+ "step": 910
627
+ },
628
+ {
629
+ "epoch": 0.36,
630
+ "learning_rate": 8.553635998392929e-05,
631
+ "loss": 0.3163,
632
+ "step": 920
633
+ },
634
+ {
635
+ "epoch": 0.36,
636
+ "learning_rate": 8.53354760948172e-05,
637
+ "loss": 0.3223,
638
+ "step": 930
639
+ },
640
+ {
641
+ "epoch": 0.36,
642
+ "learning_rate": 8.513459220570512e-05,
643
+ "loss": 0.3423,
644
+ "step": 940
645
+ },
646
+ {
647
+ "epoch": 0.37,
648
+ "learning_rate": 8.493370831659301e-05,
649
+ "loss": 0.3175,
650
+ "step": 950
651
+ },
652
+ {
653
+ "epoch": 0.37,
654
+ "learning_rate": 8.473282442748092e-05,
655
+ "loss": 0.3249,
656
+ "step": 960
657
+ },
658
+ {
659
+ "epoch": 0.37,
660
+ "learning_rate": 8.453194053836882e-05,
661
+ "loss": 0.3095,
662
+ "step": 970
663
+ },
664
+ {
665
+ "epoch": 0.38,
666
+ "learning_rate": 8.433105664925674e-05,
667
+ "loss": 0.3324,
668
+ "step": 980
669
+ },
670
+ {
671
+ "epoch": 0.38,
672
+ "learning_rate": 8.413017276014465e-05,
673
+ "loss": 0.3276,
674
+ "step": 990
675
+ },
676
+ {
677
+ "epoch": 0.39,
678
+ "learning_rate": 8.392928887103255e-05,
679
+ "loss": 0.345,
680
+ "step": 1000
681
+ },
682
+ {
683
+ "epoch": 0.39,
684
+ "eval_loss": 0.3368094265460968,
685
+ "eval_runtime": 233.1619,
686
+ "eval_samples_per_second": 2.144,
687
+ "eval_steps_per_second": 0.069,
688
+ "step": 1000
689
+ },
690
+ {
691
+ "epoch": 0.39,
692
+ "learning_rate": 8.372840498192044e-05,
693
+ "loss": 0.3389,
694
+ "step": 1010
695
+ },
696
+ {
697
+ "epoch": 0.39,
698
+ "learning_rate": 8.352752109280836e-05,
699
+ "loss": 0.321,
700
+ "step": 1020
701
+ },
702
+ {
703
+ "epoch": 0.4,
704
+ "learning_rate": 8.332663720369627e-05,
705
+ "loss": 0.3497,
706
+ "step": 1030
707
+ },
708
+ {
709
+ "epoch": 0.4,
710
+ "learning_rate": 8.312575331458418e-05,
711
+ "loss": 0.3283,
712
+ "step": 1040
713
+ },
714
+ {
715
+ "epoch": 0.41,
716
+ "learning_rate": 8.292486942547208e-05,
717
+ "loss": 0.31,
718
+ "step": 1050
719
+ },
720
+ {
721
+ "epoch": 0.41,
722
+ "learning_rate": 8.272398553635999e-05,
723
+ "loss": 0.3325,
724
+ "step": 1060
725
+ },
726
+ {
727
+ "epoch": 0.41,
728
+ "learning_rate": 8.252310164724789e-05,
729
+ "loss": 0.3103,
730
+ "step": 1070
731
+ },
732
+ {
733
+ "epoch": 0.42,
734
+ "learning_rate": 8.23222177581358e-05,
735
+ "loss": 0.3301,
736
+ "step": 1080
737
+ },
738
+ {
739
+ "epoch": 0.42,
740
+ "learning_rate": 8.21213338690237e-05,
741
+ "loss": 0.3295,
742
+ "step": 1090
743
+ },
744
+ {
745
+ "epoch": 0.42,
746
+ "learning_rate": 8.192044997991162e-05,
747
+ "loss": 0.3301,
748
+ "step": 1100
749
+ },
750
+ {
751
+ "epoch": 0.42,
752
+ "eval_loss": 0.3294503390789032,
753
+ "eval_runtime": 233.058,
754
+ "eval_samples_per_second": 2.145,
755
+ "eval_steps_per_second": 0.069,
756
+ "step": 1100
757
+ },
758
+ {
759
+ "epoch": 0.43,
760
+ "learning_rate": 8.171956609079953e-05,
761
+ "loss": 0.3269,
762
+ "step": 1110
763
+ },
764
+ {
765
+ "epoch": 0.43,
766
+ "learning_rate": 8.151868220168742e-05,
767
+ "loss": 0.347,
768
+ "step": 1120
769
+ },
770
+ {
771
+ "epoch": 0.44,
772
+ "learning_rate": 8.131779831257533e-05,
773
+ "loss": 0.3189,
774
+ "step": 1130
775
+ },
776
+ {
777
+ "epoch": 0.44,
778
+ "learning_rate": 8.111691442346325e-05,
779
+ "loss": 0.3251,
780
+ "step": 1140
781
+ },
782
+ {
783
+ "epoch": 0.44,
784
+ "learning_rate": 8.091603053435115e-05,
785
+ "loss": 0.305,
786
+ "step": 1150
787
+ },
788
+ {
789
+ "epoch": 0.45,
790
+ "learning_rate": 8.071514664523906e-05,
791
+ "loss": 0.3054,
792
+ "step": 1160
793
+ },
794
+ {
795
+ "epoch": 0.45,
796
+ "learning_rate": 8.051426275612696e-05,
797
+ "loss": 0.3426,
798
+ "step": 1170
799
+ },
800
+ {
801
+ "epoch": 0.46,
802
+ "learning_rate": 8.031337886701487e-05,
803
+ "loss": 0.3206,
804
+ "step": 1180
805
+ },
806
+ {
807
+ "epoch": 0.46,
808
+ "learning_rate": 8.011249497790278e-05,
809
+ "loss": 0.334,
810
+ "step": 1190
811
+ },
812
+ {
813
+ "epoch": 0.46,
814
+ "learning_rate": 7.991161108879068e-05,
815
+ "loss": 0.3311,
816
+ "step": 1200
817
+ },
818
+ {
819
+ "epoch": 0.46,
820
+ "eval_loss": 0.3330426812171936,
821
+ "eval_runtime": 231.4394,
822
+ "eval_samples_per_second": 2.16,
823
+ "eval_steps_per_second": 0.069,
824
+ "step": 1200
825
+ },
826
+ {
827
+ "epoch": 0.47,
828
+ "learning_rate": 7.971072719967859e-05,
829
+ "loss": 0.3326,
830
+ "step": 1210
831
+ },
832
+ {
833
+ "epoch": 0.47,
834
+ "learning_rate": 7.950984331056651e-05,
835
+ "loss": 0.3229,
836
+ "step": 1220
837
+ },
838
+ {
839
+ "epoch": 0.48,
840
+ "learning_rate": 7.93089594214544e-05,
841
+ "loss": 0.3252,
842
+ "step": 1230
843
+ },
844
+ {
845
+ "epoch": 0.48,
846
+ "learning_rate": 7.91080755323423e-05,
847
+ "loss": 0.314,
848
+ "step": 1240
849
+ },
850
+ {
851
+ "epoch": 0.48,
852
+ "learning_rate": 7.890719164323021e-05,
853
+ "loss": 0.3191,
854
+ "step": 1250
855
+ },
856
+ {
857
+ "epoch": 0.49,
858
+ "learning_rate": 7.870630775411813e-05,
859
+ "loss": 0.3193,
860
+ "step": 1260
861
+ },
862
+ {
863
+ "epoch": 0.49,
864
+ "learning_rate": 7.850542386500604e-05,
865
+ "loss": 0.3183,
866
+ "step": 1270
867
+ },
868
+ {
869
+ "epoch": 0.49,
870
+ "learning_rate": 7.830453997589394e-05,
871
+ "loss": 0.3141,
872
+ "step": 1280
873
+ },
874
+ {
875
+ "epoch": 0.5,
876
+ "learning_rate": 7.810365608678183e-05,
877
+ "loss": 0.3069,
878
+ "step": 1290
879
+ },
880
+ {
881
+ "epoch": 0.5,
882
+ "learning_rate": 7.790277219766975e-05,
883
+ "loss": 0.3203,
884
+ "step": 1300
885
+ },
886
+ {
887
+ "epoch": 0.5,
888
+ "eval_loss": 0.32666015625,
889
+ "eval_runtime": 228.195,
890
+ "eval_samples_per_second": 2.191,
891
+ "eval_steps_per_second": 0.07,
892
+ "step": 1300
893
+ },
894
+ {
895
+ "epoch": 0.51,
896
+ "learning_rate": 7.770188830855766e-05,
897
+ "loss": 0.3044,
898
+ "step": 1310
899
+ },
900
+ {
901
+ "epoch": 0.51,
902
+ "learning_rate": 7.750100441944557e-05,
903
+ "loss": 0.3052,
904
+ "step": 1320
905
+ },
906
+ {
907
+ "epoch": 0.51,
908
+ "learning_rate": 7.730012053033347e-05,
909
+ "loss": 0.3156,
910
+ "step": 1330
911
+ },
912
+ {
913
+ "epoch": 0.52,
914
+ "learning_rate": 7.709923664122138e-05,
915
+ "loss": 0.3271,
916
+ "step": 1340
917
+ },
918
+ {
919
+ "epoch": 0.52,
920
+ "learning_rate": 7.689835275210928e-05,
921
+ "loss": 0.2991,
922
+ "step": 1350
923
+ },
924
+ {
925
+ "epoch": 0.53,
926
+ "learning_rate": 7.669746886299719e-05,
927
+ "loss": 0.306,
928
+ "step": 1360
929
+ },
930
+ {
931
+ "epoch": 0.53,
932
+ "learning_rate": 7.64965849738851e-05,
933
+ "loss": 0.3122,
934
+ "step": 1370
935
+ },
936
+ {
937
+ "epoch": 0.53,
938
+ "learning_rate": 7.629570108477301e-05,
939
+ "loss": 0.3101,
940
+ "step": 1380
941
+ },
942
+ {
943
+ "epoch": 0.54,
944
+ "learning_rate": 7.609481719566092e-05,
945
+ "loss": 0.3106,
946
+ "step": 1390
947
+ },
948
+ {
949
+ "epoch": 0.54,
950
+ "learning_rate": 7.589393330654881e-05,
951
+ "loss": 0.3177,
952
+ "step": 1400
953
+ },
954
+ {
955
+ "epoch": 0.54,
956
+ "eval_loss": 0.3282645046710968,
957
+ "eval_runtime": 228.5175,
958
+ "eval_samples_per_second": 2.188,
959
+ "eval_steps_per_second": 0.07,
960
+ "step": 1400
961
+ },
962
+ {
963
+ "epoch": 0.54,
964
+ "learning_rate": 7.569304941743672e-05,
965
+ "loss": 0.2955,
966
+ "step": 1410
967
+ },
968
+ {
969
+ "epoch": 0.55,
970
+ "learning_rate": 7.549216552832464e-05,
971
+ "loss": 0.3287,
972
+ "step": 1420
973
+ },
974
+ {
975
+ "epoch": 0.55,
976
+ "learning_rate": 7.529128163921254e-05,
977
+ "loss": 0.3224,
978
+ "step": 1430
979
+ },
980
+ {
981
+ "epoch": 0.56,
982
+ "learning_rate": 7.509039775010045e-05,
983
+ "loss": 0.3231,
984
+ "step": 1440
985
+ },
986
+ {
987
+ "epoch": 0.56,
988
+ "learning_rate": 7.488951386098834e-05,
989
+ "loss": 0.3218,
990
+ "step": 1450
991
+ },
992
+ {
993
+ "epoch": 0.56,
994
+ "learning_rate": 7.468862997187626e-05,
995
+ "loss": 0.3167,
996
+ "step": 1460
997
+ },
998
+ {
999
+ "epoch": 0.57,
1000
+ "learning_rate": 7.448774608276417e-05,
1001
+ "loss": 0.3252,
1002
+ "step": 1470
1003
+ },
1004
+ {
1005
+ "epoch": 0.57,
1006
+ "learning_rate": 7.428686219365207e-05,
1007
+ "loss": 0.315,
1008
+ "step": 1480
1009
+ },
1010
+ {
1011
+ "epoch": 0.58,
1012
+ "learning_rate": 7.408597830453998e-05,
1013
+ "loss": 0.3012,
1014
+ "step": 1490
1015
+ },
1016
+ {
1017
+ "epoch": 0.58,
1018
+ "learning_rate": 7.38850944154279e-05,
1019
+ "loss": 0.3055,
1020
+ "step": 1500
1021
+ },
1022
+ {
1023
+ "epoch": 0.58,
1024
+ "eval_loss": 0.3236956000328064,
1025
+ "eval_runtime": 235.4381,
1026
+ "eval_samples_per_second": 2.124,
1027
+ "eval_steps_per_second": 0.068,
1028
+ "step": 1500
1029
+ },
1030
+ {
1031
+ "epoch": 0.58,
1032
+ "learning_rate": 7.368421052631579e-05,
1033
+ "loss": 0.3194,
1034
+ "step": 1510
1035
+ },
1036
+ {
1037
+ "epoch": 0.59,
1038
+ "learning_rate": 7.34833266372037e-05,
1039
+ "loss": 0.3254,
1040
+ "step": 1520
1041
+ },
1042
+ {
1043
+ "epoch": 0.59,
1044
+ "learning_rate": 7.32824427480916e-05,
1045
+ "loss": 0.3108,
1046
+ "step": 1530
1047
+ },
1048
+ {
1049
+ "epoch": 0.59,
1050
+ "learning_rate": 7.308155885897952e-05,
1051
+ "loss": 0.3029,
1052
+ "step": 1540
1053
+ },
1054
+ {
1055
+ "epoch": 0.6,
1056
+ "learning_rate": 7.288067496986743e-05,
1057
+ "loss": 0.319,
1058
+ "step": 1550
1059
+ },
1060
+ {
1061
+ "epoch": 0.6,
1062
+ "learning_rate": 7.267979108075532e-05,
1063
+ "loss": 0.3114,
1064
+ "step": 1560
1065
+ },
1066
+ {
1067
+ "epoch": 0.61,
1068
+ "learning_rate": 7.247890719164322e-05,
1069
+ "loss": 0.2944,
1070
+ "step": 1570
1071
+ },
1072
+ {
1073
+ "epoch": 0.61,
1074
+ "learning_rate": 7.227802330253114e-05,
1075
+ "loss": 0.3255,
1076
+ "step": 1580
1077
+ },
1078
+ {
1079
+ "epoch": 0.61,
1080
+ "learning_rate": 7.207713941341905e-05,
1081
+ "loss": 0.321,
1082
+ "step": 1590
1083
+ },
1084
+ {
1085
+ "epoch": 0.62,
1086
+ "learning_rate": 7.187625552430696e-05,
1087
+ "loss": 0.3235,
1088
+ "step": 1600
1089
+ },
1090
+ {
1091
+ "epoch": 0.62,
1092
+ "eval_loss": 0.3245675265789032,
1093
+ "eval_runtime": 229.0197,
1094
+ "eval_samples_per_second": 2.183,
1095
+ "eval_steps_per_second": 0.07,
1096
+ "step": 1600
1097
+ },
1098
+ {
1099
+ "epoch": 0.62,
1100
+ "learning_rate": 7.167537163519486e-05,
1101
+ "loss": 0.2904,
1102
+ "step": 1610
1103
+ },
1104
+ {
1105
+ "epoch": 0.63,
1106
+ "learning_rate": 7.147448774608277e-05,
1107
+ "loss": 0.3511,
1108
+ "step": 1620
1109
+ },
1110
+ {
1111
+ "epoch": 0.63,
1112
+ "learning_rate": 7.127360385697067e-05,
1113
+ "loss": 0.3088,
1114
+ "step": 1630
1115
+ },
1116
+ {
1117
+ "epoch": 0.63,
1118
+ "learning_rate": 7.107271996785858e-05,
1119
+ "loss": 0.3015,
1120
+ "step": 1640
1121
+ },
1122
+ {
1123
+ "epoch": 0.64,
1124
+ "learning_rate": 7.087183607874648e-05,
1125
+ "loss": 0.3,
1126
+ "step": 1650
1127
+ },
1128
+ {
1129
+ "epoch": 0.64,
1130
+ "learning_rate": 7.06709521896344e-05,
1131
+ "loss": 0.3487,
1132
+ "step": 1660
1133
+ },
1134
+ {
1135
+ "epoch": 0.65,
1136
+ "learning_rate": 7.04700683005223e-05,
1137
+ "loss": 0.3029,
1138
+ "step": 1670
1139
+ },
1140
+ {
1141
+ "epoch": 0.65,
1142
+ "learning_rate": 7.02691844114102e-05,
1143
+ "loss": 0.307,
1144
+ "step": 1680
1145
+ },
1146
+ {
1147
+ "epoch": 0.65,
1148
+ "learning_rate": 7.006830052229811e-05,
1149
+ "loss": 0.3026,
1150
+ "step": 1690
1151
+ },
1152
+ {
1153
+ "epoch": 0.66,
1154
+ "learning_rate": 6.986741663318603e-05,
1155
+ "loss": 0.3203,
1156
+ "step": 1700
1157
+ },
1158
+ {
1159
+ "epoch": 0.66,
1160
+ "eval_loss": 0.3210797905921936,
1161
+ "eval_runtime": 235.7603,
1162
+ "eval_samples_per_second": 2.121,
1163
+ "eval_steps_per_second": 0.068,
1164
+ "step": 1700
1165
+ },
1166
+ {
1167
+ "epoch": 0.66,
1168
+ "learning_rate": 6.966653274407393e-05,
1169
+ "loss": 0.3036,
1170
+ "step": 1710
1171
+ },
1172
+ {
1173
+ "epoch": 0.66,
1174
+ "learning_rate": 6.946564885496184e-05,
1175
+ "loss": 0.32,
1176
+ "step": 1720
1177
+ },
1178
+ {
1179
+ "epoch": 0.67,
1180
+ "learning_rate": 6.926476496584973e-05,
1181
+ "loss": 0.3065,
1182
+ "step": 1730
1183
+ },
1184
+ {
1185
+ "epoch": 0.67,
1186
+ "learning_rate": 6.906388107673765e-05,
1187
+ "loss": 0.2972,
1188
+ "step": 1740
1189
+ },
1190
+ {
1191
+ "epoch": 0.68,
1192
+ "learning_rate": 6.886299718762556e-05,
1193
+ "loss": 0.2972,
1194
+ "step": 1750
1195
+ },
1196
+ {
1197
+ "epoch": 0.68,
1198
+ "learning_rate": 6.866211329851346e-05,
1199
+ "loss": 0.321,
1200
+ "step": 1760
1201
+ },
1202
+ {
1203
+ "epoch": 0.68,
1204
+ "learning_rate": 6.846122940940137e-05,
1205
+ "loss": 0.3206,
1206
+ "step": 1770
1207
+ },
1208
+ {
1209
+ "epoch": 0.69,
1210
+ "learning_rate": 6.826034552028927e-05,
1211
+ "loss": 0.304,
1212
+ "step": 1780
1213
+ },
1214
+ {
1215
+ "epoch": 0.69,
1216
+ "learning_rate": 6.805946163117718e-05,
1217
+ "loss": 0.3101,
1218
+ "step": 1790
1219
+ },
1220
+ {
1221
+ "epoch": 0.7,
1222
+ "learning_rate": 6.785857774206509e-05,
1223
+ "loss": 0.3263,
1224
+ "step": 1800
1225
+ },
1226
+ {
1227
+ "epoch": 0.7,
1228
+ "eval_loss": 0.3179059624671936,
1229
+ "eval_runtime": 228.96,
1230
+ "eval_samples_per_second": 2.184,
1231
+ "eval_steps_per_second": 0.07,
1232
+ "step": 1800
1233
+ },
1234
+ {
1235
+ "epoch": 0.7,
1236
+ "learning_rate": 6.765769385295299e-05,
1237
+ "loss": 0.3311,
1238
+ "step": 1810
1239
+ },
1240
+ {
1241
+ "epoch": 0.7,
1242
+ "learning_rate": 6.745680996384091e-05,
1243
+ "loss": 0.3047,
1244
+ "step": 1820
1245
+ },
1246
+ {
1247
+ "epoch": 0.71,
1248
+ "learning_rate": 6.725592607472882e-05,
1249
+ "loss": 0.3085,
1250
+ "step": 1830
1251
+ },
1252
+ {
1253
+ "epoch": 0.71,
1254
+ "learning_rate": 6.705504218561671e-05,
1255
+ "loss": 0.3083,
1256
+ "step": 1840
1257
+ },
1258
+ {
1259
+ "epoch": 0.71,
1260
+ "learning_rate": 6.685415829650461e-05,
1261
+ "loss": 0.3225,
1262
+ "step": 1850
1263
+ },
1264
+ {
1265
+ "epoch": 0.72,
1266
+ "learning_rate": 6.665327440739253e-05,
1267
+ "loss": 0.3136,
1268
+ "step": 1860
1269
+ },
1270
+ {
1271
+ "epoch": 0.72,
1272
+ "learning_rate": 6.645239051828044e-05,
1273
+ "loss": 0.3191,
1274
+ "step": 1870
1275
+ },
1276
+ {
1277
+ "epoch": 0.73,
1278
+ "learning_rate": 6.625150662916835e-05,
1279
+ "loss": 0.3547,
1280
+ "step": 1880
1281
+ },
1282
+ {
1283
+ "epoch": 0.73,
1284
+ "learning_rate": 6.605062274005625e-05,
1285
+ "loss": 0.3466,
1286
+ "step": 1890
1287
+ },
1288
+ {
1289
+ "epoch": 0.73,
1290
+ "learning_rate": 6.584973885094416e-05,
1291
+ "loss": 0.2985,
1292
+ "step": 1900
1293
+ },
1294
+ {
1295
+ "epoch": 0.73,
1296
+ "eval_loss": 0.324951171875,
1297
+ "eval_runtime": 228.2301,
1298
+ "eval_samples_per_second": 2.191,
1299
+ "eval_steps_per_second": 0.07,
1300
+ "step": 1900
1301
+ },
1302
+ {
1303
+ "epoch": 0.74,
1304
+ "learning_rate": 6.564885496183206e-05,
1305
+ "loss": 0.3164,
1306
+ "step": 1910
1307
+ },
1308
+ {
1309
+ "epoch": 0.74,
1310
+ "learning_rate": 6.544797107271997e-05,
1311
+ "loss": 0.3171,
1312
+ "step": 1920
1313
+ },
1314
+ {
1315
+ "epoch": 0.75,
1316
+ "learning_rate": 6.524708718360787e-05,
1317
+ "loss": 0.2916,
1318
+ "step": 1930
1319
+ },
1320
+ {
1321
+ "epoch": 0.75,
1322
+ "learning_rate": 6.50462032944958e-05,
1323
+ "loss": 0.3082,
1324
+ "step": 1940
1325
+ },
1326
+ {
1327
+ "epoch": 0.75,
1328
+ "learning_rate": 6.484531940538369e-05,
1329
+ "loss": 0.308,
1330
+ "step": 1950
1331
+ },
1332
+ {
1333
+ "epoch": 0.76,
1334
+ "learning_rate": 6.464443551627159e-05,
1335
+ "loss": 0.3085,
1336
+ "step": 1960
1337
+ },
1338
+ {
1339
+ "epoch": 0.76,
1340
+ "learning_rate": 6.44435516271595e-05,
1341
+ "loss": 0.3062,
1342
+ "step": 1970
1343
+ },
1344
+ {
1345
+ "epoch": 0.76,
1346
+ "learning_rate": 6.424266773804742e-05,
1347
+ "loss": 0.3214,
1348
+ "step": 1980
1349
+ },
1350
+ {
1351
+ "epoch": 0.77,
1352
+ "learning_rate": 6.404178384893532e-05,
1353
+ "loss": 0.3046,
1354
+ "step": 1990
1355
+ },
1356
+ {
1357
+ "epoch": 0.77,
1358
+ "learning_rate": 6.384089995982323e-05,
1359
+ "loss": 0.3161,
1360
+ "step": 2000
1361
+ },
1362
+ {
1363
+ "epoch": 0.77,
1364
+ "eval_loss": 0.3216029703617096,
1365
+ "eval_runtime": 230.2587,
1366
+ "eval_samples_per_second": 2.171,
1367
+ "eval_steps_per_second": 0.069,
1368
+ "step": 2000
1369
+ },
1370
+ {
1371
+ "epoch": 0.78,
1372
+ "learning_rate": 6.364001607071113e-05,
1373
+ "loss": 0.3274,
1374
+ "step": 2010
1375
+ },
1376
+ {
1377
+ "epoch": 0.78,
1378
+ "learning_rate": 6.343913218159904e-05,
1379
+ "loss": 0.341,
1380
+ "step": 2020
1381
+ },
1382
+ {
1383
+ "epoch": 0.78,
1384
+ "learning_rate": 6.323824829248695e-05,
1385
+ "loss": 0.3226,
1386
+ "step": 2030
1387
+ },
1388
+ {
1389
+ "epoch": 0.79,
1390
+ "learning_rate": 6.303736440337485e-05,
1391
+ "loss": 0.3187,
1392
+ "step": 2040
1393
+ },
1394
+ {
1395
+ "epoch": 0.79,
1396
+ "learning_rate": 6.283648051426276e-05,
1397
+ "loss": 0.3003,
1398
+ "step": 2050
1399
+ },
1400
+ {
1401
+ "epoch": 0.8,
1402
+ "learning_rate": 6.263559662515066e-05,
1403
+ "loss": 0.2849,
1404
+ "step": 2060
1405
+ },
1406
+ {
1407
+ "epoch": 0.8,
1408
+ "learning_rate": 6.243471273603857e-05,
1409
+ "loss": 0.3096,
1410
+ "step": 2070
1411
+ },
1412
+ {
1413
+ "epoch": 0.8,
1414
+ "learning_rate": 6.223382884692648e-05,
1415
+ "loss": 0.2993,
1416
+ "step": 2080
1417
+ },
1418
+ {
1419
+ "epoch": 0.81,
1420
+ "learning_rate": 6.203294495781438e-05,
1421
+ "loss": 0.2984,
1422
+ "step": 2090
1423
+ },
1424
+ {
1425
+ "epoch": 0.81,
1426
+ "learning_rate": 6.18320610687023e-05,
1427
+ "loss": 0.2987,
1428
+ "step": 2100
1429
+ },
1430
+ {
1431
+ "epoch": 0.81,
1432
+ "eval_loss": 0.3201729953289032,
1433
+ "eval_runtime": 227.0048,
1434
+ "eval_samples_per_second": 2.203,
1435
+ "eval_steps_per_second": 0.07,
1436
+ "step": 2100
1437
+ },
1438
+ {
1439
+ "epoch": 0.81,
1440
+ "learning_rate": 6.16311771795902e-05,
1441
+ "loss": 0.3092,
1442
+ "step": 2110
1443
+ },
1444
+ {
1445
+ "epoch": 0.82,
1446
+ "learning_rate": 6.14302932904781e-05,
1447
+ "loss": 0.2995,
1448
+ "step": 2120
1449
+ },
1450
+ {
1451
+ "epoch": 0.82,
1452
+ "learning_rate": 6.1229409401366e-05,
1453
+ "loss": 0.3114,
1454
+ "step": 2130
1455
+ },
1456
+ {
1457
+ "epoch": 0.83,
1458
+ "learning_rate": 6.102852551225392e-05,
1459
+ "loss": 0.3077,
1460
+ "step": 2140
1461
+ },
1462
+ {
1463
+ "epoch": 0.83,
1464
+ "learning_rate": 6.082764162314183e-05,
1465
+ "loss": 0.3121,
1466
+ "step": 2150
1467
+ },
1468
+ {
1469
+ "epoch": 0.83,
1470
+ "learning_rate": 6.0626757734029736e-05,
1471
+ "loss": 0.3042,
1472
+ "step": 2160
1473
+ },
1474
+ {
1475
+ "epoch": 0.84,
1476
+ "learning_rate": 6.0425873844917635e-05,
1477
+ "loss": 0.3129,
1478
+ "step": 2170
1479
+ },
1480
+ {
1481
+ "epoch": 0.84,
1482
+ "learning_rate": 6.022498995580555e-05,
1483
+ "loss": 0.2843,
1484
+ "step": 2180
1485
+ },
1486
+ {
1487
+ "epoch": 0.85,
1488
+ "learning_rate": 6.002410606669345e-05,
1489
+ "loss": 0.2973,
1490
+ "step": 2190
1491
+ },
1492
+ {
1493
+ "epoch": 0.85,
1494
+ "learning_rate": 5.982322217758136e-05,
1495
+ "loss": 0.3405,
1496
+ "step": 2200
1497
+ },
1498
+ {
1499
+ "epoch": 0.85,
1500
+ "eval_loss": 0.3213588297367096,
1501
+ "eval_runtime": 231.1731,
1502
+ "eval_samples_per_second": 2.163,
1503
+ "eval_steps_per_second": 0.069,
1504
+ "step": 2200
1505
+ },
1506
+ {
1507
+ "epoch": 0.85,
1508
+ "learning_rate": 5.962233828846927e-05,
1509
+ "loss": 0.2974,
1510
+ "step": 2210
1511
+ },
1512
+ {
1513
+ "epoch": 0.86,
1514
+ "learning_rate": 5.942145439935718e-05,
1515
+ "loss": 0.2978,
1516
+ "step": 2220
1517
+ },
1518
+ {
1519
+ "epoch": 0.86,
1520
+ "learning_rate": 5.9220570510245076e-05,
1521
+ "loss": 0.2922,
1522
+ "step": 2230
1523
+ },
1524
+ {
1525
+ "epoch": 0.87,
1526
+ "learning_rate": 5.901968662113299e-05,
1527
+ "loss": 0.321,
1528
+ "step": 2240
1529
+ },
1530
+ {
1531
+ "epoch": 0.87,
1532
+ "learning_rate": 5.8818802732020895e-05,
1533
+ "loss": 0.2926,
1534
+ "step": 2250
1535
+ },
1536
+ {
1537
+ "epoch": 0.87,
1538
+ "learning_rate": 5.86179188429088e-05,
1539
+ "loss": 0.3175,
1540
+ "step": 2260
1541
+ },
1542
+ {
1543
+ "epoch": 0.88,
1544
+ "learning_rate": 5.841703495379671e-05,
1545
+ "loss": 0.2816,
1546
+ "step": 2270
1547
+ },
1548
+ {
1549
+ "epoch": 0.88,
1550
+ "learning_rate": 5.821615106468461e-05,
1551
+ "loss": 0.2761,
1552
+ "step": 2280
1553
+ },
1554
+ {
1555
+ "epoch": 0.88,
1556
+ "learning_rate": 5.801526717557252e-05,
1557
+ "loss": 0.2786,
1558
+ "step": 2290
1559
+ },
1560
+ {
1561
+ "epoch": 0.89,
1562
+ "learning_rate": 5.781438328646043e-05,
1563
+ "loss": 0.295,
1564
+ "step": 2300
1565
+ },
1566
+ {
1567
+ "epoch": 0.89,
1568
+ "eval_loss": 0.3176967203617096,
1569
+ "eval_runtime": 229.1487,
1570
+ "eval_samples_per_second": 2.182,
1571
+ "eval_steps_per_second": 0.07,
1572
+ "step": 2300
1573
+ },
1574
+ {
1575
+ "epoch": 0.89,
1576
+ "learning_rate": 5.7613499397348337e-05,
1577
+ "loss": 0.3299,
1578
+ "step": 2310
1579
+ },
1580
+ {
1581
+ "epoch": 0.9,
1582
+ "learning_rate": 5.741261550823624e-05,
1583
+ "loss": 0.3122,
1584
+ "step": 2320
1585
+ },
1586
+ {
1587
+ "epoch": 0.9,
1588
+ "learning_rate": 5.7211731619124155e-05,
1589
+ "loss": 0.3288,
1590
+ "step": 2330
1591
+ },
1592
+ {
1593
+ "epoch": 0.9,
1594
+ "learning_rate": 5.7010847730012054e-05,
1595
+ "loss": 0.34,
1596
+ "step": 2340
1597
+ },
1598
+ {
1599
+ "epoch": 0.91,
1600
+ "learning_rate": 5.680996384089996e-05,
1601
+ "loss": 0.3161,
1602
+ "step": 2350
1603
+ },
1604
+ {
1605
+ "epoch": 0.91,
1606
+ "learning_rate": 5.6609079951787866e-05,
1607
+ "loss": 0.2696,
1608
+ "step": 2360
1609
+ },
1610
+ {
1611
+ "epoch": 0.92,
1612
+ "learning_rate": 5.640819606267578e-05,
1613
+ "loss": 0.3285,
1614
+ "step": 2370
1615
+ },
1616
+ {
1617
+ "epoch": 0.92,
1618
+ "learning_rate": 5.6207312173563684e-05,
1619
+ "loss": 0.3158,
1620
+ "step": 2380
1621
+ },
1622
+ {
1623
+ "epoch": 0.92,
1624
+ "learning_rate": 5.600642828445158e-05,
1625
+ "loss": 0.3229,
1626
+ "step": 2390
1627
+ },
1628
+ {
1629
+ "epoch": 0.93,
1630
+ "learning_rate": 5.5805544395339496e-05,
1631
+ "loss": 0.3168,
1632
+ "step": 2400
1633
+ },
1634
+ {
1635
+ "epoch": 0.93,
1636
+ "eval_loss": 0.3182547390460968,
1637
+ "eval_runtime": 234.4281,
1638
+ "eval_samples_per_second": 2.133,
1639
+ "eval_steps_per_second": 0.068,
1640
+ "step": 2400
1641
+ },
1642
+ {
1643
+ "epoch": 0.93,
1644
+ "learning_rate": 5.56046605062274e-05,
1645
+ "loss": 0.3167,
1646
+ "step": 2410
1647
+ },
1648
+ {
1649
+ "epoch": 0.93,
1650
+ "learning_rate": 5.540377661711531e-05,
1651
+ "loss": 0.3128,
1652
+ "step": 2420
1653
+ },
1654
+ {
1655
+ "epoch": 0.94,
1656
+ "learning_rate": 5.520289272800322e-05,
1657
+ "loss": 0.293,
1658
+ "step": 2430
1659
+ },
1660
+ {
1661
+ "epoch": 0.94,
1662
+ "learning_rate": 5.5002008838891126e-05,
1663
+ "loss": 0.3113,
1664
+ "step": 2440
1665
+ },
1666
+ {
1667
+ "epoch": 0.95,
1668
+ "learning_rate": 5.4801124949779025e-05,
1669
+ "loss": 0.3167,
1670
+ "step": 2450
1671
+ },
1672
+ {
1673
+ "epoch": 0.95,
1674
+ "learning_rate": 5.460024106066694e-05,
1675
+ "loss": 0.3119,
1676
+ "step": 2460
1677
+ },
1678
+ {
1679
+ "epoch": 0.95,
1680
+ "learning_rate": 5.439935717155484e-05,
1681
+ "loss": 0.295,
1682
+ "step": 2470
1683
+ },
1684
+ {
1685
+ "epoch": 0.96,
1686
+ "learning_rate": 5.419847328244275e-05,
1687
+ "loss": 0.303,
1688
+ "step": 2480
1689
+ },
1690
+ {
1691
+ "epoch": 0.96,
1692
+ "learning_rate": 5.399758939333066e-05,
1693
+ "loss": 0.3183,
1694
+ "step": 2490
1695
+ },
1696
+ {
1697
+ "epoch": 0.97,
1698
+ "learning_rate": 5.379670550421857e-05,
1699
+ "loss": 0.3159,
1700
+ "step": 2500
1701
+ },
1702
+ {
1703
+ "epoch": 0.97,
1704
+ "eval_loss": 0.3175223171710968,
1705
+ "eval_runtime": 242.5362,
1706
+ "eval_samples_per_second": 2.062,
1707
+ "eval_steps_per_second": 0.066,
1708
+ "step": 2500
1709
+ },
1710
+ {
1711
+ "epoch": 0.97,
1712
+ "learning_rate": 5.3595821615106466e-05,
1713
+ "loss": 0.3086,
1714
+ "step": 2510
1715
+ },
1716
+ {
1717
+ "epoch": 0.97,
1718
+ "learning_rate": 5.339493772599438e-05,
1719
+ "loss": 0.3184,
1720
+ "step": 2520
1721
+ },
1722
+ {
1723
+ "epoch": 0.98,
1724
+ "learning_rate": 5.3194053836882285e-05,
1725
+ "loss": 0.3131,
1726
+ "step": 2530
1727
+ },
1728
+ {
1729
+ "epoch": 0.98,
1730
+ "learning_rate": 5.299316994777019e-05,
1731
+ "loss": 0.3099,
1732
+ "step": 2540
1733
+ },
1734
+ {
1735
+ "epoch": 0.98,
1736
+ "learning_rate": 5.27922860586581e-05,
1737
+ "loss": 0.3392,
1738
+ "step": 2550
1739
+ },
1740
+ {
1741
+ "epoch": 0.99,
1742
+ "learning_rate": 5.2591402169546e-05,
1743
+ "loss": 0.3207,
1744
+ "step": 2560
1745
+ },
1746
+ {
1747
+ "epoch": 0.99,
1748
+ "learning_rate": 5.239051828043391e-05,
1749
+ "loss": 0.307,
1750
+ "step": 2570
1751
+ },
1752
+ {
1753
+ "epoch": 1.0,
1754
+ "learning_rate": 5.218963439132182e-05,
1755
+ "loss": 0.2946,
1756
+ "step": 2580
1757
+ },
1758
+ {
1759
+ "epoch": 1.0,
1760
+ "learning_rate": 5.198875050220973e-05,
1761
+ "loss": 0.3138,
1762
+ "step": 2590
1763
+ },
1764
+ {
1765
+ "epoch": 1.0,
1766
+ "learning_rate": 5.178786661309763e-05,
1767
+ "loss": 0.2704,
1768
+ "step": 2600
1769
+ },
1770
+ {
1771
+ "epoch": 1.0,
1772
+ "eval_loss": 0.31640625,
1773
+ "eval_runtime": 234.6863,
1774
+ "eval_samples_per_second": 2.131,
1775
+ "eval_steps_per_second": 0.068,
1776
+ "step": 2600
1777
+ },
1778
+ {
1779
+ "epoch": 1.01,
1780
+ "learning_rate": 5.1586982723985545e-05,
1781
+ "loss": 0.2552,
1782
+ "step": 2610
1783
+ },
1784
+ {
1785
+ "epoch": 1.01,
1786
+ "learning_rate": 5.1386098834873444e-05,
1787
+ "loss": 0.2701,
1788
+ "step": 2620
1789
+ },
1790
+ {
1791
+ "epoch": 1.02,
1792
+ "learning_rate": 5.118521494576135e-05,
1793
+ "loss": 0.2582,
1794
+ "step": 2630
1795
+ },
1796
+ {
1797
+ "epoch": 1.02,
1798
+ "learning_rate": 5.098433105664926e-05,
1799
+ "loss": 0.2592,
1800
+ "step": 2640
1801
+ },
1802
+ {
1803
+ "epoch": 1.02,
1804
+ "learning_rate": 5.078344716753717e-05,
1805
+ "loss": 0.2804,
1806
+ "step": 2650
1807
+ },
1808
+ {
1809
+ "epoch": 1.03,
1810
+ "learning_rate": 5.0582563278425074e-05,
1811
+ "loss": 0.2553,
1812
+ "step": 2660
1813
+ },
1814
+ {
1815
+ "epoch": 1.03,
1816
+ "learning_rate": 5.038167938931297e-05,
1817
+ "loss": 0.2541,
1818
+ "step": 2670
1819
+ },
1820
+ {
1821
+ "epoch": 1.04,
1822
+ "learning_rate": 5.0180795500200886e-05,
1823
+ "loss": 0.2536,
1824
+ "step": 2680
1825
+ },
1826
+ {
1827
+ "epoch": 1.04,
1828
+ "learning_rate": 4.997991161108879e-05,
1829
+ "loss": 0.2612,
1830
+ "step": 2690
1831
+ },
1832
+ {
1833
+ "epoch": 1.04,
1834
+ "learning_rate": 4.9779027721976704e-05,
1835
+ "loss": 0.278,
1836
+ "step": 2700
1837
+ },
1838
+ {
1839
+ "epoch": 1.04,
1840
+ "eval_loss": 0.3170689046382904,
1841
+ "eval_runtime": 234.3707,
1842
+ "eval_samples_per_second": 2.133,
1843
+ "eval_steps_per_second": 0.068,
1844
+ "step": 2700
1845
+ },
1846
+ {
1847
+ "epoch": 1.05,
1848
+ "learning_rate": 4.95781438328646e-05,
1849
+ "loss": 0.2573,
1850
+ "step": 2710
1851
+ },
1852
+ {
1853
+ "epoch": 1.05,
1854
+ "learning_rate": 4.9377259943752516e-05,
1855
+ "loss": 0.2553,
1856
+ "step": 2720
1857
+ },
1858
+ {
1859
+ "epoch": 1.05,
1860
+ "learning_rate": 4.917637605464042e-05,
1861
+ "loss": 0.2675,
1862
+ "step": 2730
1863
+ },
1864
+ {
1865
+ "epoch": 1.06,
1866
+ "learning_rate": 4.897549216552833e-05,
1867
+ "loss": 0.2842,
1868
+ "step": 2740
1869
+ },
1870
+ {
1871
+ "epoch": 1.06,
1872
+ "learning_rate": 4.877460827641623e-05,
1873
+ "loss": 0.2523,
1874
+ "step": 2750
1875
+ },
1876
+ {
1877
+ "epoch": 1.07,
1878
+ "learning_rate": 4.8573724387304146e-05,
1879
+ "loss": 0.2698,
1880
+ "step": 2760
1881
+ },
1882
+ {
1883
+ "epoch": 1.07,
1884
+ "learning_rate": 4.8372840498192045e-05,
1885
+ "loss": 0.2911,
1886
+ "step": 2770
1887
+ },
1888
+ {
1889
+ "epoch": 1.07,
1890
+ "learning_rate": 4.817195660907996e-05,
1891
+ "loss": 0.2909,
1892
+ "step": 2780
1893
+ },
1894
+ {
1895
+ "epoch": 1.08,
1896
+ "learning_rate": 4.7971072719967863e-05,
1897
+ "loss": 0.2753,
1898
+ "step": 2790
1899
+ },
1900
+ {
1901
+ "epoch": 1.08,
1902
+ "learning_rate": 4.777018883085577e-05,
1903
+ "loss": 0.2703,
1904
+ "step": 2800
1905
+ },
1906
+ {
1907
+ "epoch": 1.08,
1908
+ "eval_loss": 0.312744140625,
1909
+ "eval_runtime": 232.5523,
1910
+ "eval_samples_per_second": 2.15,
1911
+ "eval_steps_per_second": 0.069,
1912
+ "step": 2800
1913
+ },
1914
+ {
1915
+ "epoch": 1.09,
1916
+ "learning_rate": 4.7569304941743675e-05,
1917
+ "loss": 0.2798,
1918
+ "step": 2810
1919
+ },
1920
+ {
1921
+ "epoch": 1.09,
1922
+ "learning_rate": 4.736842105263158e-05,
1923
+ "loss": 0.2738,
1924
+ "step": 2820
1925
+ },
1926
+ {
1927
+ "epoch": 1.09,
1928
+ "learning_rate": 4.716753716351949e-05,
1929
+ "loss": 0.2757,
1930
+ "step": 2830
1931
+ },
1932
+ {
1933
+ "epoch": 1.1,
1934
+ "learning_rate": 4.69666532744074e-05,
1935
+ "loss": 0.2441,
1936
+ "step": 2840
1937
+ },
1938
+ {
1939
+ "epoch": 1.1,
1940
+ "learning_rate": 4.67657693852953e-05,
1941
+ "loss": 0.2597,
1942
+ "step": 2850
1943
+ },
1944
+ {
1945
+ "epoch": 1.1,
1946
+ "learning_rate": 4.656488549618321e-05,
1947
+ "loss": 0.2546,
1948
+ "step": 2860
1949
+ },
1950
+ {
1951
+ "epoch": 1.11,
1952
+ "learning_rate": 4.636400160707112e-05,
1953
+ "loss": 0.2567,
1954
+ "step": 2870
1955
+ },
1956
+ {
1957
+ "epoch": 1.11,
1958
+ "learning_rate": 4.616311771795902e-05,
1959
+ "loss": 0.2677,
1960
+ "step": 2880
1961
+ },
1962
+ {
1963
+ "epoch": 1.12,
1964
+ "learning_rate": 4.596223382884693e-05,
1965
+ "loss": 0.2544,
1966
+ "step": 2890
1967
+ },
1968
+ {
1969
+ "epoch": 1.12,
1970
+ "learning_rate": 4.576134993973484e-05,
1971
+ "loss": 0.2596,
1972
+ "step": 2900
1973
+ },
1974
+ {
1975
+ "epoch": 1.12,
1976
+ "eval_loss": 0.312744140625,
1977
+ "eval_runtime": 234.4348,
1978
+ "eval_samples_per_second": 2.133,
1979
+ "eval_steps_per_second": 0.068,
1980
+ "step": 2900
1981
+ },
1982
+ {
1983
+ "epoch": 1.12,
1984
+ "learning_rate": 4.556046605062274e-05,
1985
+ "loss": 0.2678,
1986
+ "step": 2910
1987
+ },
1988
+ {
1989
+ "epoch": 1.13,
1990
+ "learning_rate": 4.535958216151065e-05,
1991
+ "loss": 0.2741,
1992
+ "step": 2920
1993
+ },
1994
+ {
1995
+ "epoch": 1.13,
1996
+ "learning_rate": 4.515869827239856e-05,
1997
+ "loss": 0.2664,
1998
+ "step": 2930
1999
+ },
2000
+ {
2001
+ "epoch": 1.14,
2002
+ "learning_rate": 4.4957814383286464e-05,
2003
+ "loss": 0.2743,
2004
+ "step": 2940
2005
+ },
2006
+ {
2007
+ "epoch": 1.14,
2008
+ "learning_rate": 4.475693049417437e-05,
2009
+ "loss": 0.2587,
2010
+ "step": 2950
2011
+ },
2012
+ {
2013
+ "epoch": 1.14,
2014
+ "learning_rate": 4.4556046605062276e-05,
2015
+ "loss": 0.2744,
2016
+ "step": 2960
2017
+ },
2018
+ {
2019
+ "epoch": 1.15,
2020
+ "learning_rate": 4.435516271595018e-05,
2021
+ "loss": 0.2637,
2022
+ "step": 2970
2023
+ },
2024
+ {
2025
+ "epoch": 1.15,
2026
+ "learning_rate": 4.4154278826838094e-05,
2027
+ "loss": 0.2604,
2028
+ "step": 2980
2029
+ },
2030
+ {
2031
+ "epoch": 1.15,
2032
+ "learning_rate": 4.395339493772599e-05,
2033
+ "loss": 0.2609,
2034
+ "step": 2990
2035
+ },
2036
+ {
2037
+ "epoch": 1.16,
2038
+ "learning_rate": 4.3752511048613906e-05,
2039
+ "loss": 0.2564,
2040
+ "step": 3000
2041
+ },
2042
+ {
2043
+ "epoch": 1.16,
2044
+ "eval_loss": 0.317138671875,
2045
+ "eval_runtime": 236.9531,
2046
+ "eval_samples_per_second": 2.11,
2047
+ "eval_steps_per_second": 0.068,
2048
+ "step": 3000
2049
+ },
2050
+ {
2051
+ "epoch": 1.16,
2052
+ "learning_rate": 4.355162715950181e-05,
2053
+ "loss": 0.2585,
2054
+ "step": 3010
2055
+ },
2056
+ {
2057
+ "epoch": 1.17,
2058
+ "learning_rate": 4.335074327038972e-05,
2059
+ "loss": 0.272,
2060
+ "step": 3020
2061
+ },
2062
+ {
2063
+ "epoch": 1.17,
2064
+ "learning_rate": 4.3149859381277623e-05,
2065
+ "loss": 0.2593,
2066
+ "step": 3030
2067
+ },
2068
+ {
2069
+ "epoch": 1.17,
2070
+ "learning_rate": 4.2948975492165536e-05,
2071
+ "loss": 0.2658,
2072
+ "step": 3040
2073
+ },
2074
+ {
2075
+ "epoch": 1.18,
2076
+ "learning_rate": 4.2748091603053435e-05,
2077
+ "loss": 0.2643,
2078
+ "step": 3050
2079
+ },
2080
+ {
2081
+ "epoch": 1.18,
2082
+ "learning_rate": 4.254720771394135e-05,
2083
+ "loss": 0.2631,
2084
+ "step": 3060
2085
+ },
2086
+ {
2087
+ "epoch": 1.19,
2088
+ "learning_rate": 4.234632382482925e-05,
2089
+ "loss": 0.2376,
2090
+ "step": 3070
2091
+ },
2092
+ {
2093
+ "epoch": 1.19,
2094
+ "learning_rate": 4.214543993571716e-05,
2095
+ "loss": 0.2645,
2096
+ "step": 3080
2097
+ },
2098
+ {
2099
+ "epoch": 1.19,
2100
+ "learning_rate": 4.1944556046605065e-05,
2101
+ "loss": 0.2903,
2102
+ "step": 3090
2103
+ },
2104
+ {
2105
+ "epoch": 1.2,
2106
+ "learning_rate": 4.174367215749297e-05,
2107
+ "loss": 0.2502,
2108
+ "step": 3100
2109
+ },
2110
+ {
2111
+ "epoch": 1.2,
2112
+ "eval_loss": 0.3148367702960968,
2113
+ "eval_runtime": 232.416,
2114
+ "eval_samples_per_second": 2.151,
2115
+ "eval_steps_per_second": 0.069,
2116
+ "step": 3100
2117
+ },
2118
+ {
2119
+ "epoch": 1.2,
2120
+ "learning_rate": 4.154278826838088e-05,
2121
+ "loss": 0.2389,
2122
+ "step": 3110
2123
+ },
2124
+ {
2125
+ "epoch": 1.21,
2126
+ "learning_rate": 4.134190437926879e-05,
2127
+ "loss": 0.2575,
2128
+ "step": 3120
2129
+ },
2130
+ {
2131
+ "epoch": 1.21,
2132
+ "learning_rate": 4.114102049015669e-05,
2133
+ "loss": 0.2825,
2134
+ "step": 3130
2135
+ },
2136
+ {
2137
+ "epoch": 1.21,
2138
+ "learning_rate": 4.09401366010446e-05,
2139
+ "loss": 0.2574,
2140
+ "step": 3140
2141
+ },
2142
+ {
2143
+ "epoch": 1.22,
2144
+ "learning_rate": 4.073925271193251e-05,
2145
+ "loss": 0.2707,
2146
+ "step": 3150
2147
+ },
2148
+ {
2149
+ "epoch": 1.22,
2150
+ "learning_rate": 4.053836882282041e-05,
2151
+ "loss": 0.2589,
2152
+ "step": 3160
2153
+ },
2154
+ {
2155
+ "epoch": 1.22,
2156
+ "learning_rate": 4.033748493370832e-05,
2157
+ "loss": 0.2646,
2158
+ "step": 3170
2159
+ },
2160
+ {
2161
+ "epoch": 1.23,
2162
+ "learning_rate": 4.0136601044596224e-05,
2163
+ "loss": 0.2685,
2164
+ "step": 3180
2165
+ },
2166
+ {
2167
+ "epoch": 1.23,
2168
+ "learning_rate": 3.993571715548413e-05,
2169
+ "loss": 0.2595,
2170
+ "step": 3190
2171
+ },
2172
+ {
2173
+ "epoch": 1.24,
2174
+ "learning_rate": 3.973483326637204e-05,
2175
+ "loss": 0.2477,
2176
+ "step": 3200
2177
+ },
2178
+ {
2179
+ "epoch": 1.24,
2180
+ "eval_loss": 0.3164760172367096,
2181
+ "eval_runtime": 229.5059,
2182
+ "eval_samples_per_second": 2.179,
2183
+ "eval_steps_per_second": 0.07,
2184
+ "step": 3200
2185
+ },
2186
+ {
2187
+ "epoch": 1.24,
2188
+ "learning_rate": 3.953394937725994e-05,
2189
+ "loss": 0.2598,
2190
+ "step": 3210
2191
+ },
2192
+ {
2193
+ "epoch": 1.24,
2194
+ "learning_rate": 3.9333065488147854e-05,
2195
+ "loss": 0.2626,
2196
+ "step": 3220
2197
+ },
2198
+ {
2199
+ "epoch": 1.25,
2200
+ "learning_rate": 3.913218159903576e-05,
2201
+ "loss": 0.2538,
2202
+ "step": 3230
2203
+ },
2204
+ {
2205
+ "epoch": 1.25,
2206
+ "learning_rate": 3.8931297709923666e-05,
2207
+ "loss": 0.2562,
2208
+ "step": 3240
2209
+ },
2210
+ {
2211
+ "epoch": 1.26,
2212
+ "learning_rate": 3.873041382081157e-05,
2213
+ "loss": 0.2625,
2214
+ "step": 3250
2215
+ },
2216
+ {
2217
+ "epoch": 1.26,
2218
+ "learning_rate": 3.8529529931699484e-05,
2219
+ "loss": 0.2638,
2220
+ "step": 3260
2221
+ },
2222
+ {
2223
+ "epoch": 1.26,
2224
+ "learning_rate": 3.8328646042587384e-05,
2225
+ "loss": 0.2415,
2226
+ "step": 3270
2227
+ },
2228
+ {
2229
+ "epoch": 1.27,
2230
+ "learning_rate": 3.8127762153475296e-05,
2231
+ "loss": 0.2894,
2232
+ "step": 3280
2233
+ },
2234
+ {
2235
+ "epoch": 1.27,
2236
+ "learning_rate": 3.79268782643632e-05,
2237
+ "loss": 0.2606,
2238
+ "step": 3290
2239
+ },
2240
+ {
2241
+ "epoch": 1.27,
2242
+ "learning_rate": 3.772599437525111e-05,
2243
+ "loss": 0.2607,
2244
+ "step": 3300
2245
+ },
2246
+ {
2247
+ "epoch": 1.27,
2248
+ "eval_loss": 0.3150809109210968,
2249
+ "eval_runtime": 231.3186,
2250
+ "eval_samples_per_second": 2.162,
2251
+ "eval_steps_per_second": 0.069,
2252
+ "step": 3300
2253
+ },
2254
+ {
2255
+ "epoch": 1.28,
2256
+ "learning_rate": 3.7525110486139014e-05,
2257
+ "loss": 0.2853,
2258
+ "step": 3310
2259
+ },
2260
+ {
2261
+ "epoch": 1.28,
2262
+ "learning_rate": 3.732422659702692e-05,
2263
+ "loss": 0.2597,
2264
+ "step": 3320
2265
+ },
2266
+ {
2267
+ "epoch": 1.29,
2268
+ "learning_rate": 3.7123342707914825e-05,
2269
+ "loss": 0.2653,
2270
+ "step": 3330
2271
+ },
2272
+ {
2273
+ "epoch": 1.29,
2274
+ "learning_rate": 3.692245881880274e-05,
2275
+ "loss": 0.2626,
2276
+ "step": 3340
2277
+ },
2278
+ {
2279
+ "epoch": 1.29,
2280
+ "learning_rate": 3.672157492969064e-05,
2281
+ "loss": 0.2465,
2282
+ "step": 3350
2283
+ },
2284
+ {
2285
+ "epoch": 1.3,
2286
+ "learning_rate": 3.652069104057855e-05,
2287
+ "loss": 0.2997,
2288
+ "step": 3360
2289
+ },
2290
+ {
2291
+ "epoch": 1.3,
2292
+ "learning_rate": 3.6319807151466455e-05,
2293
+ "loss": 0.2614,
2294
+ "step": 3370
2295
+ },
2296
+ {
2297
+ "epoch": 1.31,
2298
+ "learning_rate": 3.611892326235436e-05,
2299
+ "loss": 0.2708,
2300
+ "step": 3380
2301
+ },
2302
+ {
2303
+ "epoch": 1.31,
2304
+ "learning_rate": 3.591803937324227e-05,
2305
+ "loss": 0.2633,
2306
+ "step": 3390
2307
+ },
2308
+ {
2309
+ "epoch": 1.31,
2310
+ "learning_rate": 3.571715548413018e-05,
2311
+ "loss": 0.2607,
2312
+ "step": 3400
2313
+ },
2314
+ {
2315
+ "epoch": 1.31,
2316
+ "eval_loss": 0.309326171875,
2317
+ "eval_runtime": 234.7514,
2318
+ "eval_samples_per_second": 2.13,
2319
+ "eval_steps_per_second": 0.068,
2320
+ "step": 3400
2321
+ },
2322
+ {
2323
+ "epoch": 1.32,
2324
+ "learning_rate": 3.551627159501808e-05,
2325
+ "loss": 0.2605,
2326
+ "step": 3410
2327
+ },
2328
+ {
2329
+ "epoch": 1.32,
2330
+ "learning_rate": 3.531538770590599e-05,
2331
+ "loss": 0.2557,
2332
+ "step": 3420
2333
+ },
2334
+ {
2335
+ "epoch": 1.32,
2336
+ "learning_rate": 3.511450381679389e-05,
2337
+ "loss": 0.2566,
2338
+ "step": 3430
2339
+ },
2340
+ {
2341
+ "epoch": 1.33,
2342
+ "learning_rate": 3.49136199276818e-05,
2343
+ "loss": 0.2502,
2344
+ "step": 3440
2345
+ },
2346
+ {
2347
+ "epoch": 1.33,
2348
+ "learning_rate": 3.471273603856971e-05,
2349
+ "loss": 0.2512,
2350
+ "step": 3450
2351
+ },
2352
+ {
2353
+ "epoch": 1.34,
2354
+ "learning_rate": 3.4511852149457614e-05,
2355
+ "loss": 0.2966,
2356
+ "step": 3460
2357
+ },
2358
+ {
2359
+ "epoch": 1.34,
2360
+ "learning_rate": 3.431096826034552e-05,
2361
+ "loss": 0.2408,
2362
+ "step": 3470
2363
+ },
2364
+ {
2365
+ "epoch": 1.34,
2366
+ "learning_rate": 3.411008437123343e-05,
2367
+ "loss": 0.2667,
2368
+ "step": 3480
2369
+ },
2370
+ {
2371
+ "epoch": 1.35,
2372
+ "learning_rate": 3.390920048212133e-05,
2373
+ "loss": 0.2851,
2374
+ "step": 3490
2375
+ },
2376
+ {
2377
+ "epoch": 1.35,
2378
+ "learning_rate": 3.3708316593009245e-05,
2379
+ "loss": 0.2488,
2380
+ "step": 3500
2381
+ },
2382
+ {
2383
+ "epoch": 1.35,
2384
+ "eval_loss": 0.3107561469078064,
2385
+ "eval_runtime": 232.6614,
2386
+ "eval_samples_per_second": 2.149,
2387
+ "eval_steps_per_second": 0.069,
2388
+ "step": 3500
2389
+ },
2390
+ {
2391
+ "epoch": 1.36,
2392
+ "learning_rate": 3.350743270389715e-05,
2393
+ "loss": 0.2942,
2394
+ "step": 3510
2395
+ },
2396
+ {
2397
+ "epoch": 1.36,
2398
+ "learning_rate": 3.3306548814785056e-05,
2399
+ "loss": 0.2605,
2400
+ "step": 3520
2401
+ },
2402
+ {
2403
+ "epoch": 1.36,
2404
+ "learning_rate": 3.310566492567296e-05,
2405
+ "loss": 0.2352,
2406
+ "step": 3530
2407
+ },
2408
+ {
2409
+ "epoch": 1.37,
2410
+ "learning_rate": 3.290478103656087e-05,
2411
+ "loss": 0.2371,
2412
+ "step": 3540
2413
+ },
2414
+ {
2415
+ "epoch": 1.37,
2416
+ "learning_rate": 3.2703897147448774e-05,
2417
+ "loss": 0.2597,
2418
+ "step": 3550
2419
+ },
2420
+ {
2421
+ "epoch": 1.38,
2422
+ "learning_rate": 3.2503013258336686e-05,
2423
+ "loss": 0.2452,
2424
+ "step": 3560
2425
+ },
2426
+ {
2427
+ "epoch": 1.38,
2428
+ "learning_rate": 3.2302129369224585e-05,
2429
+ "loss": 0.2617,
2430
+ "step": 3570
2431
+ },
2432
+ {
2433
+ "epoch": 1.38,
2434
+ "learning_rate": 3.21012454801125e-05,
2435
+ "loss": 0.2664,
2436
+ "step": 3580
2437
+ },
2438
+ {
2439
+ "epoch": 1.39,
2440
+ "learning_rate": 3.1900361591000404e-05,
2441
+ "loss": 0.28,
2442
+ "step": 3590
2443
+ },
2444
+ {
2445
+ "epoch": 1.39,
2446
+ "learning_rate": 3.169947770188831e-05,
2447
+ "loss": 0.2376,
2448
+ "step": 3600
2449
+ },
2450
+ {
2451
+ "epoch": 1.39,
2452
+ "eval_loss": 0.3105817437171936,
2453
+ "eval_runtime": 232.1998,
2454
+ "eval_samples_per_second": 2.153,
2455
+ "eval_steps_per_second": 0.069,
2456
+ "step": 3600
2457
+ },
2458
+ {
2459
+ "epoch": 1.39,
2460
+ "learning_rate": 3.1498593812776215e-05,
2461
+ "loss": 0.2667,
2462
+ "step": 3610
2463
+ },
2464
+ {
2465
+ "epoch": 1.4,
2466
+ "learning_rate": 3.129770992366413e-05,
2467
+ "loss": 0.2599,
2468
+ "step": 3620
2469
+ },
2470
+ {
2471
+ "epoch": 1.4,
2472
+ "learning_rate": 3.109682603455203e-05,
2473
+ "loss": 0.2656,
2474
+ "step": 3630
2475
+ },
2476
+ {
2477
+ "epoch": 1.41,
2478
+ "learning_rate": 3.089594214543994e-05,
2479
+ "loss": 0.2311,
2480
+ "step": 3640
2481
+ },
2482
+ {
2483
+ "epoch": 1.41,
2484
+ "learning_rate": 3.0695058256327845e-05,
2485
+ "loss": 0.2506,
2486
+ "step": 3650
2487
+ },
2488
+ {
2489
+ "epoch": 1.41,
2490
+ "learning_rate": 3.049417436721575e-05,
2491
+ "loss": 0.2598,
2492
+ "step": 3660
2493
+ },
2494
+ {
2495
+ "epoch": 1.42,
2496
+ "learning_rate": 3.0293290478103657e-05,
2497
+ "loss": 0.2654,
2498
+ "step": 3670
2499
+ },
2500
+ {
2501
+ "epoch": 1.42,
2502
+ "learning_rate": 3.0092406588991563e-05,
2503
+ "loss": 0.2454,
2504
+ "step": 3680
2505
+ },
2506
+ {
2507
+ "epoch": 1.43,
2508
+ "learning_rate": 2.9891522699879472e-05,
2509
+ "loss": 0.2513,
2510
+ "step": 3690
2511
+ },
2512
+ {
2513
+ "epoch": 1.43,
2514
+ "learning_rate": 2.9690638810767378e-05,
2515
+ "loss": 0.2642,
2516
+ "step": 3700
2517
+ },
2518
+ {
2519
+ "epoch": 1.43,
2520
+ "eval_loss": 0.3145228922367096,
2521
+ "eval_runtime": 240.0973,
2522
+ "eval_samples_per_second": 2.082,
2523
+ "eval_steps_per_second": 0.067,
2524
+ "step": 3700
2525
+ },
2526
+ {
2527
+ "epoch": 1.43,
2528
+ "learning_rate": 2.9489754921655284e-05,
2529
+ "loss": 0.2386,
2530
+ "step": 3710
2531
+ },
2532
+ {
2533
+ "epoch": 1.44,
2534
+ "learning_rate": 2.9288871032543193e-05,
2535
+ "loss": 0.2449,
2536
+ "step": 3720
2537
+ },
2538
+ {
2539
+ "epoch": 1.44,
2540
+ "learning_rate": 2.90879871434311e-05,
2541
+ "loss": 0.2581,
2542
+ "step": 3730
2543
+ },
2544
+ {
2545
+ "epoch": 1.44,
2546
+ "learning_rate": 2.8887103254319005e-05,
2547
+ "loss": 0.25,
2548
+ "step": 3740
2549
+ },
2550
+ {
2551
+ "epoch": 1.45,
2552
+ "learning_rate": 2.8686219365206914e-05,
2553
+ "loss": 0.2552,
2554
+ "step": 3750
2555
+ },
2556
+ {
2557
+ "epoch": 1.45,
2558
+ "learning_rate": 2.848533547609482e-05,
2559
+ "loss": 0.2819,
2560
+ "step": 3760
2561
+ },
2562
+ {
2563
+ "epoch": 1.46,
2564
+ "learning_rate": 2.8284451586982725e-05,
2565
+ "loss": 0.2491,
2566
+ "step": 3770
2567
+ },
2568
+ {
2569
+ "epoch": 1.46,
2570
+ "learning_rate": 2.8083567697870635e-05,
2571
+ "loss": 0.2582,
2572
+ "step": 3780
2573
+ },
2574
+ {
2575
+ "epoch": 1.46,
2576
+ "learning_rate": 2.7882683808758537e-05,
2577
+ "loss": 0.2479,
2578
+ "step": 3790
2579
+ },
2580
+ {
2581
+ "epoch": 1.47,
2582
+ "learning_rate": 2.7681799919646446e-05,
2583
+ "loss": 0.2686,
2584
+ "step": 3800
2585
+ },
2586
+ {
2587
+ "epoch": 1.47,
2588
+ "eval_loss": 0.3105817437171936,
2589
+ "eval_runtime": 229.2876,
2590
+ "eval_samples_per_second": 2.181,
2591
+ "eval_steps_per_second": 0.07,
2592
+ "step": 3800
2593
+ },
2594
+ {
2595
+ "epoch": 1.47,
2596
+ "learning_rate": 2.7480916030534355e-05,
2597
+ "loss": 0.265,
2598
+ "step": 3810
2599
+ },
2600
+ {
2601
+ "epoch": 1.48,
2602
+ "learning_rate": 2.7280032141422258e-05,
2603
+ "loss": 0.2557,
2604
+ "step": 3820
2605
+ },
2606
+ {
2607
+ "epoch": 1.48,
2608
+ "learning_rate": 2.7079148252310167e-05,
2609
+ "loss": 0.2458,
2610
+ "step": 3830
2611
+ },
2612
+ {
2613
+ "epoch": 1.48,
2614
+ "learning_rate": 2.6878264363198073e-05,
2615
+ "loss": 0.2683,
2616
+ "step": 3840
2617
+ },
2618
+ {
2619
+ "epoch": 1.49,
2620
+ "learning_rate": 2.667738047408598e-05,
2621
+ "loss": 0.2494,
2622
+ "step": 3850
2623
+ },
2624
+ {
2625
+ "epoch": 1.49,
2626
+ "learning_rate": 2.6476496584973888e-05,
2627
+ "loss": 0.2631,
2628
+ "step": 3860
2629
+ },
2630
+ {
2631
+ "epoch": 1.49,
2632
+ "learning_rate": 2.6275612695861794e-05,
2633
+ "loss": 0.256,
2634
+ "step": 3870
2635
+ },
2636
+ {
2637
+ "epoch": 1.5,
2638
+ "learning_rate": 2.60747288067497e-05,
2639
+ "loss": 0.2686,
2640
+ "step": 3880
2641
+ },
2642
+ {
2643
+ "epoch": 1.5,
2644
+ "learning_rate": 2.587384491763761e-05,
2645
+ "loss": 0.2562,
2646
+ "step": 3890
2647
+ },
2648
+ {
2649
+ "epoch": 1.51,
2650
+ "learning_rate": 2.5672961028525515e-05,
2651
+ "loss": 0.2474,
2652
+ "step": 3900
2653
+ },
2654
+ {
2655
+ "epoch": 1.51,
2656
+ "eval_loss": 0.3085588812828064,
2657
+ "eval_runtime": 234.5548,
2658
+ "eval_samples_per_second": 2.132,
2659
+ "eval_steps_per_second": 0.068,
2660
+ "step": 3900
2661
+ },
2662
+ {
2663
+ "epoch": 1.51,
2664
+ "learning_rate": 2.547207713941342e-05,
2665
+ "loss": 0.2615,
2666
+ "step": 3910
2667
+ },
2668
+ {
2669
+ "epoch": 1.51,
2670
+ "learning_rate": 2.527119325030133e-05,
2671
+ "loss": 0.2588,
2672
+ "step": 3920
2673
+ },
2674
+ {
2675
+ "epoch": 1.52,
2676
+ "learning_rate": 2.5070309361189232e-05,
2677
+ "loss": 0.2585,
2678
+ "step": 3930
2679
+ },
2680
+ {
2681
+ "epoch": 1.52,
2682
+ "learning_rate": 2.486942547207714e-05,
2683
+ "loss": 0.2703,
2684
+ "step": 3940
2685
+ },
2686
+ {
2687
+ "epoch": 1.53,
2688
+ "learning_rate": 2.4668541582965047e-05,
2689
+ "loss": 0.2341,
2690
+ "step": 3950
2691
+ },
2692
+ {
2693
+ "epoch": 1.53,
2694
+ "learning_rate": 2.4467657693852956e-05,
2695
+ "loss": 0.2556,
2696
+ "step": 3960
2697
+ },
2698
+ {
2699
+ "epoch": 1.53,
2700
+ "learning_rate": 2.4266773804740862e-05,
2701
+ "loss": 0.2621,
2702
+ "step": 3970
2703
+ },
2704
+ {
2705
+ "epoch": 1.54,
2706
+ "learning_rate": 2.4065889915628768e-05,
2707
+ "loss": 0.2614,
2708
+ "step": 3980
2709
+ },
2710
+ {
2711
+ "epoch": 1.54,
2712
+ "learning_rate": 2.3865006026516677e-05,
2713
+ "loss": 0.2531,
2714
+ "step": 3990
2715
+ },
2716
+ {
2717
+ "epoch": 1.54,
2718
+ "learning_rate": 2.3664122137404583e-05,
2719
+ "loss": 0.265,
2720
+ "step": 4000
2721
+ },
2722
+ {
2723
+ "epoch": 1.54,
2724
+ "eval_loss": 0.3104073703289032,
2725
+ "eval_runtime": 234.8987,
2726
+ "eval_samples_per_second": 2.129,
2727
+ "eval_steps_per_second": 0.068,
2728
+ "step": 4000
2729
+ },
2730
+ {
2731
+ "epoch": 1.55,
2732
+ "learning_rate": 2.346323824829249e-05,
2733
+ "loss": 0.2684,
2734
+ "step": 4010
2735
+ },
2736
+ {
2737
+ "epoch": 1.55,
2738
+ "learning_rate": 2.3262354359180395e-05,
2739
+ "loss": 0.2399,
2740
+ "step": 4020
2741
+ },
2742
+ {
2743
+ "epoch": 1.56,
2744
+ "learning_rate": 2.3061470470068304e-05,
2745
+ "loss": 0.2632,
2746
+ "step": 4030
2747
+ },
2748
+ {
2749
+ "epoch": 1.56,
2750
+ "learning_rate": 2.286058658095621e-05,
2751
+ "loss": 0.2659,
2752
+ "step": 4040
2753
+ },
2754
+ {
2755
+ "epoch": 1.56,
2756
+ "learning_rate": 2.2659702691844116e-05,
2757
+ "loss": 0.2662,
2758
+ "step": 4050
2759
+ },
2760
+ {
2761
+ "epoch": 1.57,
2762
+ "learning_rate": 2.245881880273202e-05,
2763
+ "loss": 0.24,
2764
+ "step": 4060
2765
+ },
2766
+ {
2767
+ "epoch": 1.57,
2768
+ "learning_rate": 2.225793491361993e-05,
2769
+ "loss": 0.248,
2770
+ "step": 4070
2771
+ },
2772
+ {
2773
+ "epoch": 1.58,
2774
+ "learning_rate": 2.2057051024507836e-05,
2775
+ "loss": 0.2675,
2776
+ "step": 4080
2777
+ },
2778
+ {
2779
+ "epoch": 1.58,
2780
+ "learning_rate": 2.1856167135395742e-05,
2781
+ "loss": 0.263,
2782
+ "step": 4090
2783
+ },
2784
+ {
2785
+ "epoch": 1.58,
2786
+ "learning_rate": 2.165528324628365e-05,
2787
+ "loss": 0.259,
2788
+ "step": 4100
2789
+ },
2790
+ {
2791
+ "epoch": 1.58,
2792
+ "eval_loss": 0.3070940375328064,
2793
+ "eval_runtime": 232.4523,
2794
+ "eval_samples_per_second": 2.151,
2795
+ "eval_steps_per_second": 0.069,
2796
+ "step": 4100
2797
+ },
2798
+ {
2799
+ "epoch": 1.59,
2800
+ "learning_rate": 2.1454399357171557e-05,
2801
+ "loss": 0.2708,
2802
+ "step": 4110
2803
+ },
2804
+ {
2805
+ "epoch": 1.59,
2806
+ "learning_rate": 2.1253515468059463e-05,
2807
+ "loss": 0.2591,
2808
+ "step": 4120
2809
+ },
2810
+ {
2811
+ "epoch": 1.6,
2812
+ "learning_rate": 2.105263157894737e-05,
2813
+ "loss": 0.2451,
2814
+ "step": 4130
2815
+ },
2816
+ {
2817
+ "epoch": 1.6,
2818
+ "learning_rate": 2.0851747689835278e-05,
2819
+ "loss": 0.2636,
2820
+ "step": 4140
2821
+ },
2822
+ {
2823
+ "epoch": 1.6,
2824
+ "learning_rate": 2.0650863800723184e-05,
2825
+ "loss": 0.2754,
2826
+ "step": 4150
2827
+ },
2828
+ {
2829
+ "epoch": 1.61,
2830
+ "learning_rate": 2.044997991161109e-05,
2831
+ "loss": 0.2639,
2832
+ "step": 4160
2833
+ },
2834
+ {
2835
+ "epoch": 1.61,
2836
+ "learning_rate": 2.0249096022499e-05,
2837
+ "loss": 0.258,
2838
+ "step": 4170
2839
+ },
2840
+ {
2841
+ "epoch": 1.61,
2842
+ "learning_rate": 2.0048212133386905e-05,
2843
+ "loss": 0.2585,
2844
+ "step": 4180
2845
+ },
2846
+ {
2847
+ "epoch": 1.62,
2848
+ "learning_rate": 1.984732824427481e-05,
2849
+ "loss": 0.2735,
2850
+ "step": 4190
2851
+ },
2852
+ {
2853
+ "epoch": 1.62,
2854
+ "learning_rate": 1.9646444355162716e-05,
2855
+ "loss": 0.2529,
2856
+ "step": 4200
2857
+ },
2858
+ {
2859
+ "epoch": 1.62,
2860
+ "eval_loss": 0.3050014078617096,
2861
+ "eval_runtime": 231.7346,
2862
+ "eval_samples_per_second": 2.158,
2863
+ "eval_steps_per_second": 0.069,
2864
+ "step": 4200
2865
+ },
2866
+ {
2867
+ "epoch": 1.63,
2868
+ "learning_rate": 1.9445560466050626e-05,
2869
+ "loss": 0.2392,
2870
+ "step": 4210
2871
+ },
2872
+ {
2873
+ "epoch": 1.63,
2874
+ "learning_rate": 1.924467657693853e-05,
2875
+ "loss": 0.2495,
2876
+ "step": 4220
2877
+ },
2878
+ {
2879
+ "epoch": 1.63,
2880
+ "learning_rate": 1.9043792687826437e-05,
2881
+ "loss": 0.2435,
2882
+ "step": 4230
2883
+ },
2884
+ {
2885
+ "epoch": 1.64,
2886
+ "learning_rate": 1.8842908798714343e-05,
2887
+ "loss": 0.2352,
2888
+ "step": 4240
2889
+ },
2890
+ {
2891
+ "epoch": 1.64,
2892
+ "learning_rate": 1.8642024909602252e-05,
2893
+ "loss": 0.2506,
2894
+ "step": 4250
2895
+ },
2896
+ {
2897
+ "epoch": 1.65,
2898
+ "learning_rate": 1.8441141020490158e-05,
2899
+ "loss": 0.2384,
2900
+ "step": 4260
2901
+ },
2902
+ {
2903
+ "epoch": 1.65,
2904
+ "learning_rate": 1.8240257131378064e-05,
2905
+ "loss": 0.256,
2906
+ "step": 4270
2907
+ },
2908
+ {
2909
+ "epoch": 1.65,
2910
+ "learning_rate": 1.8039373242265973e-05,
2911
+ "loss": 0.255,
2912
+ "step": 4280
2913
+ },
2914
+ {
2915
+ "epoch": 1.66,
2916
+ "learning_rate": 1.783848935315388e-05,
2917
+ "loss": 0.2645,
2918
+ "step": 4290
2919
+ },
2920
+ {
2921
+ "epoch": 1.66,
2922
+ "learning_rate": 1.7637605464041785e-05,
2923
+ "loss": 0.2624,
2924
+ "step": 4300
2925
+ },
2926
+ {
2927
+ "epoch": 1.66,
2928
+ "eval_loss": 0.308837890625,
2929
+ "eval_runtime": 236.1172,
2930
+ "eval_samples_per_second": 2.118,
2931
+ "eval_steps_per_second": 0.068,
2932
+ "step": 4300
2933
+ },
2934
+ {
2935
+ "epoch": 1.66,
2936
+ "learning_rate": 1.743672157492969e-05,
2937
+ "loss": 0.2557,
2938
+ "step": 4310
2939
+ },
2940
+ {
2941
+ "epoch": 1.67,
2942
+ "learning_rate": 1.72358376858176e-05,
2943
+ "loss": 0.2391,
2944
+ "step": 4320
2945
+ },
2946
+ {
2947
+ "epoch": 1.67,
2948
+ "learning_rate": 1.7034953796705506e-05,
2949
+ "loss": 0.2361,
2950
+ "step": 4330
2951
+ },
2952
+ {
2953
+ "epoch": 1.68,
2954
+ "learning_rate": 1.683406990759341e-05,
2955
+ "loss": 0.2389,
2956
+ "step": 4340
2957
+ },
2958
+ {
2959
+ "epoch": 1.68,
2960
+ "learning_rate": 1.663318601848132e-05,
2961
+ "loss": 0.2491,
2962
+ "step": 4350
2963
+ },
2964
+ {
2965
+ "epoch": 1.68,
2966
+ "learning_rate": 1.6432302129369227e-05,
2967
+ "loss": 0.2556,
2968
+ "step": 4360
2969
+ },
2970
+ {
2971
+ "epoch": 1.69,
2972
+ "learning_rate": 1.6231418240257132e-05,
2973
+ "loss": 0.2963,
2974
+ "step": 4370
2975
+ },
2976
+ {
2977
+ "epoch": 1.69,
2978
+ "learning_rate": 1.6030534351145038e-05,
2979
+ "loss": 0.2571,
2980
+ "step": 4380
2981
+ },
2982
+ {
2983
+ "epoch": 1.7,
2984
+ "learning_rate": 1.5829650462032947e-05,
2985
+ "loss": 0.2707,
2986
+ "step": 4390
2987
+ },
2988
+ {
2989
+ "epoch": 1.7,
2990
+ "learning_rate": 1.5628766572920853e-05,
2991
+ "loss": 0.2513,
2992
+ "step": 4400
2993
+ },
2994
+ {
2995
+ "epoch": 1.7,
2996
+ "eval_loss": 0.3080357015132904,
2997
+ "eval_runtime": 231.5394,
2998
+ "eval_samples_per_second": 2.159,
2999
+ "eval_steps_per_second": 0.069,
3000
+ "step": 4400
3001
+ },
3002
+ {
3003
+ "epoch": 1.7,
3004
+ "learning_rate": 1.542788268380876e-05,
3005
+ "loss": 0.2542,
3006
+ "step": 4410
3007
+ },
3008
+ {
3009
+ "epoch": 1.71,
3010
+ "learning_rate": 1.5226998794696665e-05,
3011
+ "loss": 0.2576,
3012
+ "step": 4420
3013
+ },
3014
+ {
3015
+ "epoch": 1.71,
3016
+ "learning_rate": 1.5026114905584574e-05,
3017
+ "loss": 0.2314,
3018
+ "step": 4430
3019
+ },
3020
+ {
3021
+ "epoch": 1.71,
3022
+ "learning_rate": 1.482523101647248e-05,
3023
+ "loss": 0.242,
3024
+ "step": 4440
3025
+ },
3026
+ {
3027
+ "epoch": 1.72,
3028
+ "learning_rate": 1.4624347127360386e-05,
3029
+ "loss": 0.2599,
3030
+ "step": 4450
3031
+ },
3032
+ {
3033
+ "epoch": 1.72,
3034
+ "learning_rate": 1.4423463238248295e-05,
3035
+ "loss": 0.267,
3036
+ "step": 4460
3037
+ },
3038
+ {
3039
+ "epoch": 1.73,
3040
+ "learning_rate": 1.42225793491362e-05,
3041
+ "loss": 0.2508,
3042
+ "step": 4470
3043
+ },
3044
+ {
3045
+ "epoch": 1.73,
3046
+ "learning_rate": 1.4021695460024107e-05,
3047
+ "loss": 0.2845,
3048
+ "step": 4480
3049
+ },
3050
+ {
3051
+ "epoch": 1.73,
3052
+ "learning_rate": 1.3820811570912012e-05,
3053
+ "loss": 0.2647,
3054
+ "step": 4490
3055
+ },
3056
+ {
3057
+ "epoch": 1.74,
3058
+ "learning_rate": 1.3619927681799922e-05,
3059
+ "loss": 0.2667,
3060
+ "step": 4500
3061
+ },
3062
+ {
3063
+ "epoch": 1.74,
3064
+ "eval_loss": 0.3072335422039032,
3065
+ "eval_runtime": 231.7522,
3066
+ "eval_samples_per_second": 2.157,
3067
+ "eval_steps_per_second": 0.069,
3068
+ "step": 4500
3069
+ },
3070
+ {
3071
+ "epoch": 1.74,
3072
+ "learning_rate": 1.3419043792687827e-05,
3073
+ "loss": 0.2759,
3074
+ "step": 4510
3075
+ },
3076
+ {
3077
+ "epoch": 1.75,
3078
+ "learning_rate": 1.3218159903575733e-05,
3079
+ "loss": 0.2618,
3080
+ "step": 4520
3081
+ },
3082
+ {
3083
+ "epoch": 1.75,
3084
+ "learning_rate": 1.3017276014463642e-05,
3085
+ "loss": 0.2465,
3086
+ "step": 4530
3087
+ },
3088
+ {
3089
+ "epoch": 1.75,
3090
+ "learning_rate": 1.2816392125351548e-05,
3091
+ "loss": 0.2601,
3092
+ "step": 4540
3093
+ },
3094
+ {
3095
+ "epoch": 1.76,
3096
+ "learning_rate": 1.2615508236239454e-05,
3097
+ "loss": 0.2567,
3098
+ "step": 4550
3099
+ },
3100
+ {
3101
+ "epoch": 1.76,
3102
+ "learning_rate": 1.2414624347127362e-05,
3103
+ "loss": 0.2366,
3104
+ "step": 4560
3105
+ },
3106
+ {
3107
+ "epoch": 1.77,
3108
+ "learning_rate": 1.2213740458015267e-05,
3109
+ "loss": 0.2563,
3110
+ "step": 4570
3111
+ },
3112
+ {
3113
+ "epoch": 1.77,
3114
+ "learning_rate": 1.2012856568903175e-05,
3115
+ "loss": 0.2599,
3116
+ "step": 4580
3117
+ },
3118
+ {
3119
+ "epoch": 1.77,
3120
+ "learning_rate": 1.181197267979108e-05,
3121
+ "loss": 0.2677,
3122
+ "step": 4590
3123
+ },
3124
+ {
3125
+ "epoch": 1.78,
3126
+ "learning_rate": 1.1611088790678988e-05,
3127
+ "loss": 0.2559,
3128
+ "step": 4600
3129
+ },
3130
+ {
3131
+ "epoch": 1.78,
3132
+ "eval_loss": 0.3053501546382904,
3133
+ "eval_runtime": 227.8046,
3134
+ "eval_samples_per_second": 2.195,
3135
+ "eval_steps_per_second": 0.07,
3136
+ "step": 4600
3137
+ },
3138
+ {
3139
+ "epoch": 1.78,
3140
+ "learning_rate": 1.1410204901566896e-05,
3141
+ "loss": 0.2499,
3142
+ "step": 4610
3143
+ },
3144
+ {
3145
+ "epoch": 1.78,
3146
+ "learning_rate": 1.1209321012454802e-05,
3147
+ "loss": 0.27,
3148
+ "step": 4620
3149
+ },
3150
+ {
3151
+ "epoch": 1.79,
3152
+ "learning_rate": 1.1008437123342709e-05,
3153
+ "loss": 0.2661,
3154
+ "step": 4630
3155
+ },
3156
+ {
3157
+ "epoch": 1.79,
3158
+ "learning_rate": 1.0807553234230615e-05,
3159
+ "loss": 0.2456,
3160
+ "step": 4640
3161
+ },
3162
+ {
3163
+ "epoch": 1.8,
3164
+ "learning_rate": 1.0606669345118522e-05,
3165
+ "loss": 0.2711,
3166
+ "step": 4650
3167
+ },
3168
+ {
3169
+ "epoch": 1.8,
3170
+ "learning_rate": 1.0405785456006428e-05,
3171
+ "loss": 0.2593,
3172
+ "step": 4660
3173
+ },
3174
+ {
3175
+ "epoch": 1.8,
3176
+ "learning_rate": 1.0204901566894336e-05,
3177
+ "loss": 0.2536,
3178
+ "step": 4670
3179
+ },
3180
+ {
3181
+ "epoch": 1.81,
3182
+ "learning_rate": 1.0004017677782242e-05,
3183
+ "loss": 0.2514,
3184
+ "step": 4680
3185
+ },
3186
+ {
3187
+ "epoch": 1.81,
3188
+ "learning_rate": 9.803133788670149e-06,
3189
+ "loss": 0.2489,
3190
+ "step": 4690
3191
+ },
3192
+ {
3193
+ "epoch": 1.82,
3194
+ "learning_rate": 9.602249899558057e-06,
3195
+ "loss": 0.2238,
3196
+ "step": 4700
3197
+ },
3198
+ {
3199
+ "epoch": 1.82,
3200
+ "eval_loss": 0.3055942952632904,
3201
+ "eval_runtime": 229.6273,
3202
+ "eval_samples_per_second": 2.177,
3203
+ "eval_steps_per_second": 0.07,
3204
+ "step": 4700
3205
+ },
3206
+ {
3207
+ "epoch": 1.82,
3208
+ "learning_rate": 9.401366010445962e-06,
3209
+ "loss": 0.2482,
3210
+ "step": 4710
3211
+ },
3212
+ {
3213
+ "epoch": 1.82,
3214
+ "learning_rate": 9.20048212133387e-06,
3215
+ "loss": 0.2443,
3216
+ "step": 4720
3217
+ },
3218
+ {
3219
+ "epoch": 1.83,
3220
+ "learning_rate": 8.999598232221776e-06,
3221
+ "loss": 0.2386,
3222
+ "step": 4730
3223
+ },
3224
+ {
3225
+ "epoch": 1.83,
3226
+ "learning_rate": 8.798714343109683e-06,
3227
+ "loss": 0.2502,
3228
+ "step": 4740
3229
+ },
3230
+ {
3231
+ "epoch": 1.83,
3232
+ "learning_rate": 8.597830453997589e-06,
3233
+ "loss": 0.2552,
3234
+ "step": 4750
3235
+ },
3236
+ {
3237
+ "epoch": 1.84,
3238
+ "learning_rate": 8.396946564885497e-06,
3239
+ "loss": 0.2473,
3240
+ "step": 4760
3241
+ },
3242
+ {
3243
+ "epoch": 1.84,
3244
+ "learning_rate": 8.196062675773402e-06,
3245
+ "loss": 0.249,
3246
+ "step": 4770
3247
+ },
3248
+ {
3249
+ "epoch": 1.85,
3250
+ "learning_rate": 7.99517878666131e-06,
3251
+ "loss": 0.2373,
3252
+ "step": 4780
3253
+ },
3254
+ {
3255
+ "epoch": 1.85,
3256
+ "learning_rate": 7.794294897549218e-06,
3257
+ "loss": 0.2211,
3258
+ "step": 4790
3259
+ },
3260
+ {
3261
+ "epoch": 1.85,
3262
+ "learning_rate": 7.593411008437123e-06,
3263
+ "loss": 0.2686,
3264
+ "step": 4800
3265
+ },
3266
+ {
3267
+ "epoch": 1.85,
3268
+ "eval_loss": 0.3053850531578064,
3269
+ "eval_runtime": 231.0424,
3270
+ "eval_samples_per_second": 2.164,
3271
+ "eval_steps_per_second": 0.069,
3272
+ "step": 4800
3273
+ },
3274
+ {
3275
+ "epoch": 1.86,
3276
+ "learning_rate": 7.392527119325031e-06,
3277
+ "loss": 0.2641,
3278
+ "step": 4810
3279
+ },
3280
+ {
3281
+ "epoch": 1.86,
3282
+ "learning_rate": 7.191643230212937e-06,
3283
+ "loss": 0.2389,
3284
+ "step": 4820
3285
+ },
3286
+ {
3287
+ "epoch": 1.87,
3288
+ "learning_rate": 6.990759341100844e-06,
3289
+ "loss": 0.2573,
3290
+ "step": 4830
3291
+ },
3292
+ {
3293
+ "epoch": 1.87,
3294
+ "learning_rate": 6.78987545198875e-06,
3295
+ "loss": 0.2335,
3296
+ "step": 4840
3297
+ },
3298
+ {
3299
+ "epoch": 1.87,
3300
+ "learning_rate": 6.5889915628766575e-06,
3301
+ "loss": 0.2724,
3302
+ "step": 4850
3303
+ },
3304
+ {
3305
+ "epoch": 1.88,
3306
+ "learning_rate": 6.388107673764563e-06,
3307
+ "loss": 0.2552,
3308
+ "step": 4860
3309
+ },
3310
+ {
3311
+ "epoch": 1.88,
3312
+ "learning_rate": 6.187223784652471e-06,
3313
+ "loss": 0.2422,
3314
+ "step": 4870
3315
+ },
3316
+ {
3317
+ "epoch": 1.88,
3318
+ "learning_rate": 5.9863398955403775e-06,
3319
+ "loss": 0.252,
3320
+ "step": 4880
3321
+ },
3322
+ {
3323
+ "epoch": 1.89,
3324
+ "learning_rate": 5.785456006428284e-06,
3325
+ "loss": 0.2619,
3326
+ "step": 4890
3327
+ },
3328
+ {
3329
+ "epoch": 1.89,
3330
+ "learning_rate": 5.584572117316191e-06,
3331
+ "loss": 0.2684,
3332
+ "step": 4900
3333
+ },
3334
+ {
3335
+ "epoch": 1.89,
3336
+ "eval_loss": 0.3049665093421936,
3337
+ "eval_runtime": 231.0605,
3338
+ "eval_samples_per_second": 2.164,
3339
+ "eval_steps_per_second": 0.069,
3340
+ "step": 4900
3341
+ },
3342
+ {
3343
+ "epoch": 1.9,
3344
+ "learning_rate": 5.383688228204098e-06,
3345
+ "loss": 0.2503,
3346
+ "step": 4910
3347
+ },
3348
+ {
3349
+ "epoch": 1.9,
3350
+ "learning_rate": 5.182804339092005e-06,
3351
+ "loss": 0.2581,
3352
+ "step": 4920
3353
+ },
3354
+ {
3355
+ "epoch": 1.9,
3356
+ "learning_rate": 4.981920449979912e-06,
3357
+ "loss": 0.2445,
3358
+ "step": 4930
3359
+ },
3360
+ {
3361
+ "epoch": 1.91,
3362
+ "learning_rate": 4.781036560867818e-06,
3363
+ "loss": 0.2527,
3364
+ "step": 4940
3365
+ },
3366
+ {
3367
+ "epoch": 1.91,
3368
+ "learning_rate": 4.580152671755725e-06,
3369
+ "loss": 0.2579,
3370
+ "step": 4950
3371
+ },
3372
+ {
3373
+ "epoch": 1.92,
3374
+ "learning_rate": 4.379268782643632e-06,
3375
+ "loss": 0.2558,
3376
+ "step": 4960
3377
+ },
3378
+ {
3379
+ "epoch": 1.92,
3380
+ "learning_rate": 4.178384893531538e-06,
3381
+ "loss": 0.2598,
3382
+ "step": 4970
3383
+ },
3384
+ {
3385
+ "epoch": 1.92,
3386
+ "learning_rate": 3.977501004419446e-06,
3387
+ "loss": 0.231,
3388
+ "step": 4980
3389
+ },
3390
+ {
3391
+ "epoch": 1.93,
3392
+ "learning_rate": 3.776617115307352e-06,
3393
+ "loss": 0.2665,
3394
+ "step": 4990
3395
+ },
3396
+ {
3397
+ "epoch": 1.93,
3398
+ "learning_rate": 3.5757332261952597e-06,
3399
+ "loss": 0.247,
3400
+ "step": 5000
3401
+ },
3402
+ {
3403
+ "epoch": 1.93,
3404
+ "eval_loss": 0.3040597140789032,
3405
+ "eval_runtime": 233.9026,
3406
+ "eval_samples_per_second": 2.138,
3407
+ "eval_steps_per_second": 0.068,
3408
+ "step": 5000
3409
+ },
3410
+ {
3411
+ "epoch": 1.94,
3412
+ "learning_rate": 3.3748493370831664e-06,
3413
+ "loss": 0.2499,
3414
+ "step": 5010
3415
+ },
3416
+ {
3417
+ "epoch": 1.94,
3418
+ "learning_rate": 3.173965447971073e-06,
3419
+ "loss": 0.241,
3420
+ "step": 5020
3421
+ },
3422
+ {
3423
+ "epoch": 1.94,
3424
+ "learning_rate": 2.9730815588589797e-06,
3425
+ "loss": 0.2455,
3426
+ "step": 5030
3427
+ },
3428
+ {
3429
+ "epoch": 1.95,
3430
+ "learning_rate": 2.7721976697468864e-06,
3431
+ "loss": 0.2592,
3432
+ "step": 5040
3433
+ },
3434
+ {
3435
+ "epoch": 1.95,
3436
+ "learning_rate": 2.571313780634793e-06,
3437
+ "loss": 0.2829,
3438
+ "step": 5050
3439
+ },
3440
+ {
3441
+ "epoch": 1.95,
3442
+ "learning_rate": 2.3704298915227e-06,
3443
+ "loss": 0.251,
3444
+ "step": 5060
3445
+ },
3446
+ {
3447
+ "epoch": 1.96,
3448
+ "learning_rate": 2.169546002410607e-06,
3449
+ "loss": 0.2587,
3450
+ "step": 5070
3451
+ },
3452
+ {
3453
+ "epoch": 1.96,
3454
+ "learning_rate": 1.9686621132985135e-06,
3455
+ "loss": 0.2574,
3456
+ "step": 5080
3457
+ },
3458
+ {
3459
+ "epoch": 1.97,
3460
+ "learning_rate": 1.7677782241864203e-06,
3461
+ "loss": 0.2697,
3462
+ "step": 5090
3463
+ },
3464
+ {
3465
+ "epoch": 1.97,
3466
+ "learning_rate": 1.566894335074327e-06,
3467
+ "loss": 0.2477,
3468
+ "step": 5100
3469
+ },
3470
+ {
3471
+ "epoch": 1.97,
3472
+ "eval_loss": 0.303466796875,
3473
+ "eval_runtime": 231.3482,
3474
+ "eval_samples_per_second": 2.161,
3475
+ "eval_steps_per_second": 0.069,
3476
+ "step": 5100
3477
+ },
3478
+ {
3479
+ "epoch": 1.97,
3480
+ "learning_rate": 1.3660104459622339e-06,
3481
+ "loss": 0.2642,
3482
+ "step": 5110
3483
+ },
3484
+ {
3485
+ "epoch": 1.98,
3486
+ "learning_rate": 1.1651265568501408e-06,
3487
+ "loss": 0.2616,
3488
+ "step": 5120
3489
+ },
3490
+ {
3491
+ "epoch": 1.98,
3492
+ "learning_rate": 9.642426677380474e-07,
3493
+ "loss": 0.2268,
3494
+ "step": 5130
3495
+ },
3496
+ {
3497
+ "epoch": 1.99,
3498
+ "learning_rate": 7.633587786259542e-07,
3499
+ "loss": 0.2561,
3500
+ "step": 5140
3501
+ },
3502
+ {
3503
+ "epoch": 1.99,
3504
+ "learning_rate": 5.62474889513861e-07,
3505
+ "loss": 0.2544,
3506
+ "step": 5150
3507
+ },
3508
+ {
3509
+ "epoch": 1.99,
3510
+ "learning_rate": 3.6159100040176776e-07,
3511
+ "loss": 0.2465,
3512
+ "step": 5160
3513
+ },
3514
+ {
3515
+ "epoch": 2.0,
3516
+ "learning_rate": 1.6070711128967456e-07,
3517
+ "loss": 0.2721,
3518
+ "step": 5170
3519
+ }
3520
+ ],
3521
+ "max_steps": 5178,
3522
+ "num_train_epochs": 2,
3523
+ "total_flos": 3.628920535154426e+18,
3524
+ "trial_name": null,
3525
+ "trial_params": null
3526
+ }
owl-con/checkpoint-5178/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:966f65be3c927ae7b3f355c69d036bd4d1faf28b059ed8e0b3f7caf80b643f7f
3
+ size 3707
owl-con/params.txt ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ adam_beta1: 0.9
2
+ adam_beta2: 0.999
3
+ adam_eps: 1e-08
4
+ all_params: True
5
+ bf16: True
6
+ clip_grad: 1.0
7
+ ddp_find_unused_parameters: False
8
+ do_train: True
9
+ eval_iters: 100
10
+ finetuned_ckpt: None
11
+ gradient_accumulation_steps: 1
12
+ gradient_checkpointing: True
13
+ inference_mode: False
14
+ local_rank: 0
15
+ logging_nan_inf_filter: False
16
+ lora_alpha: 32
17
+ lora_dropout: 0.05
18
+ lora_r: 32
19
+ loss_objective: sequential
20
+ lr: 0.0001
21
+ micro_batch_size: 16
22
+ min_lr: 1e-07
23
+ mm_config: configs/video_mix.yaml
24
+ num_training_steps: 4236
25
+ num_warmup_steps: 200
26
+ num_workers: 32
27
+ pretrained_ckpt: /local2/hbansal/video_text/mplug-owl-llama-7b-video
28
+ save_interval: None
29
+ save_path: /local2/hbansal/video_text/mplugowl_wipeout_data_second_stage_lora_all_params_32_sequential_mix_1e-4/
30
+ seq_length: 256
31
+ train_epochs: 2
32
+ train_visual_abstractor: False
33
+ use_lora: True
34
+ use_qv: False
35
+ wandb_run_name: mplugowl_wipeout_data_second_stage_lora_all_params_32_sequential_mix_1e-4
36
+ weight_decay: 0.0001