Hugging Face's logo

Neelectric
/
OLMo-2-1124-7B-Instruct_SFTv02.03

Text Generation
Transformers
Safetensors
olmo2
Generated from Trainer
open-r1
trl
sft
conversational
Model card Files
xet
Community
Neelectric commited on
Commit
a6e3d1a
·
verified ·
1 Parent(s): 47a9594

Model save

Browse files
Files changed (8) hide show
  1. README.md +58 -0
  2. all_results.json +8 -0
  3. generation_config.json +6 -0
  4. model-00001-of-00003.safetensors +1 -1
  5. model-00002-of-00003.safetensors +1 -1
  6. model-00003-of-00003.safetensors +1 -1
  7. train_results.json +8 -0
  8. trainer_state.json +1275 -0
README.md ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: allenai/OLMo-2-1124-7B-Instruct
3
+ library_name: transformers
4
+ model_name: OLMo-2-1124-7B-Instruct_SFTv02.03
5
+ tags:
6
+ - generated_from_trainer
7
+ - trl
8
+ - sft
9
+ licence: license
10
+ ---
11
+
12
+ # Model Card for OLMo-2-1124-7B-Instruct_SFTv02.03
13
+
14
+ This model is a fine-tuned version of [allenai/OLMo-2-1124-7B-Instruct](https://huggingface.co/allenai/OLMo-2-1124-7B-Instruct).
15
+ It has been trained using [TRL](https://github.com/huggingface/trl).
16
+
17
+ ## Quick start
18
+
19
+ ```python
20
+ from transformers import pipeline
21
+
22
+ question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
23
+ generator = pipeline("text-generation", model="Neelectric/OLMo-2-1124-7B-Instruct_SFTv02.03", device="cuda")
24
+ output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
25
+ print(output["generated_text"])
26
+ ```
27
+
28
+ ## Training procedure
29
+
30
+ [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/neelectric/open-r1_SFT/runs/6s8lmdqe)
31
+
32
+
33
+ This model was trained with SFT.
34
+
35
+ ### Framework versions
36
+
37
+ - TRL: 0.17.0
38
+ - Transformers: 4.51.3
39
+ - Pytorch: 2.6.0
40
+ - Datasets: 3.5.1
41
+ - Tokenizers: 0.21.1
42
+
43
+ ## Citations
44
+
45
+
46
+
47
+ Cite TRL as:
48
+
49
+ ```bibtex
50
+ @misc{vonwerra2022trl,
51
+ title = {{TRL: Transformer Reinforcement Learning}},
52
+ author = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallou{\'e}dec},
53
+ year = 2020,
54
+ journal = {GitHub repository},
55
+ publisher = {GitHub},
56
+ howpublished = {\url{https://github.com/huggingface/trl}}
57
+ }
58
+ ```
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "total_flos": 6.182777521931551e+18,
3
+ "train_loss": 0.6460230693027571,
4
+ "train_runtime": 6408.7951,
5
+ "train_samples": 69132,
6
+ "train_samples_per_second": 10.787,
7
+ "train_steps_per_second": 0.024
8
+ }
generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "eos_token_id": 100257,
4
+ "pad_token_id": 100277,
5
+ "transformers_version": "4.51.3"
6
+ }
model-00001-of-00003.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4418c58238b722a13f1610362882e5a8eacd731c09ff882e3eca422c48b94bbf
3
  size 4970591184
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99d6c1defd45e0968020e537e5a40ca42ae55aadde504de4334427e1c342806e
3
  size 4970591184
model-00002-of-00003.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:bc4e3ae5ee848391967bfdee5bd0fc7da47b6a765e70ab3db4c37d3f1f0989bd
3
  size 4981161496
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c45dbc126ef58fe7e85083842c21f921ace4f13007d9ff223d492b42f2dc8593
3
  size 4981161496
model-00003-of-00003.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e3429689c90cb376bab318c11737e9b45cd10472bc105d8d3fd2d01f3866af56
3
  size 4645523448
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cefee0f1e8e0258abe596c5f3363a17795b8d3c933d6732abcbedbfe14ee7d6e
3
  size 4645523448
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "total_flos": 6.182777521931551e+18,
3
+ "train_loss": 0.6460230693027571,
4
+ "train_runtime": 6408.7951,
5
+ "train_samples": 69132,
6
+ "train_samples_per_second": 10.787,
7
+ "train_steps_per_second": 0.024
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1275 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.9967637540453075,
6
+ "eval_steps": 500,
7
+ "global_step": 154,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.006472491909385114,
14
+ "grad_norm": 5.306360721588135,
15
+ "learning_rate": 0.0,
16
+ "loss": 1.1501,
17
+ "num_tokens": 963559.0,
18
+ "step": 1
19
+ },
20
+ {
21
+ "epoch": 0.012944983818770227,
22
+ "grad_norm": 5.30866003036499,
23
+ "learning_rate": 6.25e-08,
24
+ "loss": 1.1455,
25
+ "num_tokens": 1927760.0,
26
+ "step": 2
27
+ },
28
+ {
29
+ "epoch": 0.019417475728155338,
30
+ "grad_norm": 5.259037971496582,
31
+ "learning_rate": 1.25e-07,
32
+ "loss": 1.1393,
33
+ "num_tokens": 2896369.0,
34
+ "step": 3
35
+ },
36
+ {
37
+ "epoch": 0.025889967637540454,
38
+ "grad_norm": 5.178048610687256,
39
+ "learning_rate": 1.875e-07,
40
+ "loss": 1.133,
41
+ "num_tokens": 3886583.0,
42
+ "step": 4
43
+ },
44
+ {
45
+ "epoch": 0.032362459546925564,
46
+ "grad_norm": 5.211574554443359,
47
+ "learning_rate": 2.5e-07,
48
+ "loss": 1.1368,
49
+ "num_tokens": 4864476.0,
50
+ "step": 5
51
+ },
52
+ {
53
+ "epoch": 0.038834951456310676,
54
+ "grad_norm": 5.201828479766846,
55
+ "learning_rate": 3.1249999999999997e-07,
56
+ "loss": 1.1455,
57
+ "num_tokens": 5873295.0,
58
+ "step": 6
59
+ },
60
+ {
61
+ "epoch": 0.045307443365695796,
62
+ "grad_norm": 5.131554126739502,
63
+ "learning_rate": 3.75e-07,
64
+ "loss": 1.1259,
65
+ "num_tokens": 6876484.0,
66
+ "step": 7
67
+ },
68
+ {
69
+ "epoch": 0.05177993527508091,
70
+ "grad_norm": 5.083103179931641,
71
+ "learning_rate": 4.375e-07,
72
+ "loss": 1.1167,
73
+ "num_tokens": 7871664.0,
74
+ "step": 8
75
+ },
76
+ {
77
+ "epoch": 0.05825242718446602,
78
+ "grad_norm": 5.106949806213379,
79
+ "learning_rate": 5e-07,
80
+ "loss": 1.1324,
81
+ "num_tokens": 8869931.0,
82
+ "step": 9
83
+ },
84
+ {
85
+ "epoch": 0.06472491909385113,
86
+ "grad_norm": 4.861895561218262,
87
+ "learning_rate": 5.625e-07,
88
+ "loss": 1.1082,
89
+ "num_tokens": 9874687.0,
90
+ "step": 10
91
+ },
92
+ {
93
+ "epoch": 0.07119741100323625,
94
+ "grad_norm": 4.900995254516602,
95
+ "learning_rate": 6.249999999999999e-07,
96
+ "loss": 1.1035,
97
+ "num_tokens": 10842628.0,
98
+ "step": 11
99
+ },
100
+ {
101
+ "epoch": 0.07766990291262135,
102
+ "grad_norm": 4.93906307220459,
103
+ "learning_rate": 6.875e-07,
104
+ "loss": 1.1207,
105
+ "num_tokens": 11815438.0,
106
+ "step": 12
107
+ },
108
+ {
109
+ "epoch": 0.08414239482200647,
110
+ "grad_norm": 4.3877153396606445,
111
+ "learning_rate": 7.5e-07,
112
+ "loss": 1.0708,
113
+ "num_tokens": 12788222.0,
114
+ "step": 13
115
+ },
116
+ {
117
+ "epoch": 0.09061488673139159,
118
+ "grad_norm": 4.357248306274414,
119
+ "learning_rate": 8.125e-07,
120
+ "loss": 1.0679,
121
+ "num_tokens": 13756270.0,
122
+ "step": 14
123
+ },
124
+ {
125
+ "epoch": 0.0970873786407767,
126
+ "grad_norm": 4.3989644050598145,
127
+ "learning_rate": 8.75e-07,
128
+ "loss": 1.068,
129
+ "num_tokens": 14689813.0,
130
+ "step": 15
131
+ },
132
+ {
133
+ "epoch": 0.10355987055016182,
134
+ "grad_norm": 4.206644535064697,
135
+ "learning_rate": 9.374999999999999e-07,
136
+ "loss": 1.0545,
137
+ "num_tokens": 15670657.0,
138
+ "step": 16
139
+ },
140
+ {
141
+ "epoch": 0.11003236245954692,
142
+ "grad_norm": 4.120600700378418,
143
+ "learning_rate": 1e-06,
144
+ "loss": 1.0357,
145
+ "num_tokens": 16640491.0,
146
+ "step": 17
147
+ },
148
+ {
149
+ "epoch": 0.11650485436893204,
150
+ "grad_norm": 3.066406011581421,
151
+ "learning_rate": 1e-06,
152
+ "loss": 0.9571,
153
+ "num_tokens": 17611388.0,
154
+ "step": 18
155
+ },
156
+ {
157
+ "epoch": 0.12297734627831715,
158
+ "grad_norm": 3.0390701293945312,
159
+ "learning_rate": 1e-06,
160
+ "loss": 0.9557,
161
+ "num_tokens": 18582653.0,
162
+ "step": 19
163
+ },
164
+ {
165
+ "epoch": 0.12944983818770225,
166
+ "grad_norm": 2.996507406234741,
167
+ "learning_rate": 1e-06,
168
+ "loss": 0.9467,
169
+ "num_tokens": 19557459.0,
170
+ "step": 20
171
+ },
172
+ {
173
+ "epoch": 0.13592233009708737,
174
+ "grad_norm": 2.852745771408081,
175
+ "learning_rate": 1e-06,
176
+ "loss": 0.9414,
177
+ "num_tokens": 20524158.0,
178
+ "step": 21
179
+ },
180
+ {
181
+ "epoch": 0.1423948220064725,
182
+ "grad_norm": 2.7312402725219727,
183
+ "learning_rate": 1e-06,
184
+ "loss": 0.9161,
185
+ "num_tokens": 21502674.0,
186
+ "step": 22
187
+ },
188
+ {
189
+ "epoch": 0.1488673139158576,
190
+ "grad_norm": 2.712592124938965,
191
+ "learning_rate": 1e-06,
192
+ "loss": 0.9201,
193
+ "num_tokens": 22446294.0,
194
+ "step": 23
195
+ },
196
+ {
197
+ "epoch": 0.1553398058252427,
198
+ "grad_norm": 2.51646089553833,
199
+ "learning_rate": 1e-06,
200
+ "loss": 0.8971,
201
+ "num_tokens": 23430225.0,
202
+ "step": 24
203
+ },
204
+ {
205
+ "epoch": 0.16181229773462782,
206
+ "grad_norm": 1.9695550203323364,
207
+ "learning_rate": 1e-06,
208
+ "loss": 0.8188,
209
+ "num_tokens": 24380419.0,
210
+ "step": 25
211
+ },
212
+ {
213
+ "epoch": 0.16828478964401294,
214
+ "grad_norm": 1.7356574535369873,
215
+ "learning_rate": 1e-06,
216
+ "loss": 0.7798,
217
+ "num_tokens": 25343499.0,
218
+ "step": 26
219
+ },
220
+ {
221
+ "epoch": 0.17475728155339806,
222
+ "grad_norm": 1.5638344287872314,
223
+ "learning_rate": 1e-06,
224
+ "loss": 0.7747,
225
+ "num_tokens": 26315921.0,
226
+ "step": 27
227
+ },
228
+ {
229
+ "epoch": 0.18122977346278318,
230
+ "grad_norm": 1.524434208869934,
231
+ "learning_rate": 1e-06,
232
+ "loss": 0.7787,
233
+ "num_tokens": 27255686.0,
234
+ "step": 28
235
+ },
236
+ {
237
+ "epoch": 0.18770226537216828,
238
+ "grad_norm": 1.3938937187194824,
239
+ "learning_rate": 1e-06,
240
+ "loss": 0.7647,
241
+ "num_tokens": 28232449.0,
242
+ "step": 29
243
+ },
244
+ {
245
+ "epoch": 0.1941747572815534,
246
+ "grad_norm": 1.3155360221862793,
247
+ "learning_rate": 1e-06,
248
+ "loss": 0.7544,
249
+ "num_tokens": 29213099.0,
250
+ "step": 30
251
+ },
252
+ {
253
+ "epoch": 0.20064724919093851,
254
+ "grad_norm": 1.1851674318313599,
255
+ "learning_rate": 1e-06,
256
+ "loss": 0.751,
257
+ "num_tokens": 30210546.0,
258
+ "step": 31
259
+ },
260
+ {
261
+ "epoch": 0.20711974110032363,
262
+ "grad_norm": 1.0744044780731201,
263
+ "learning_rate": 1e-06,
264
+ "loss": 0.7319,
265
+ "num_tokens": 31190215.0,
266
+ "step": 32
267
+ },
268
+ {
269
+ "epoch": 0.21359223300970873,
270
+ "grad_norm": 1.027134895324707,
271
+ "learning_rate": 1e-06,
272
+ "loss": 0.7286,
273
+ "num_tokens": 32152368.0,
274
+ "step": 33
275
+ },
276
+ {
277
+ "epoch": 0.22006472491909385,
278
+ "grad_norm": 0.9877621531486511,
279
+ "learning_rate": 1e-06,
280
+ "loss": 0.7382,
281
+ "num_tokens": 33138410.0,
282
+ "step": 34
283
+ },
284
+ {
285
+ "epoch": 0.22653721682847897,
286
+ "grad_norm": 0.9485065937042236,
287
+ "learning_rate": 1e-06,
288
+ "loss": 0.7276,
289
+ "num_tokens": 34103559.0,
290
+ "step": 35
291
+ },
292
+ {
293
+ "epoch": 0.23300970873786409,
294
+ "grad_norm": 0.8876718282699585,
295
+ "learning_rate": 1e-06,
296
+ "loss": 0.7243,
297
+ "num_tokens": 35099702.0,
298
+ "step": 36
299
+ },
300
+ {
301
+ "epoch": 0.23948220064724918,
302
+ "grad_norm": 0.781726598739624,
303
+ "learning_rate": 1e-06,
304
+ "loss": 0.715,
305
+ "num_tokens": 36067246.0,
306
+ "step": 37
307
+ },
308
+ {
309
+ "epoch": 0.2459546925566343,
310
+ "grad_norm": 0.6435908675193787,
311
+ "learning_rate": 1e-06,
312
+ "loss": 0.6923,
313
+ "num_tokens": 37074157.0,
314
+ "step": 38
315
+ },
316
+ {
317
+ "epoch": 0.2524271844660194,
318
+ "grad_norm": 0.5644338130950928,
319
+ "learning_rate": 1e-06,
320
+ "loss": 0.6791,
321
+ "num_tokens": 38026864.0,
322
+ "step": 39
323
+ },
324
+ {
325
+ "epoch": 0.2588996763754045,
326
+ "grad_norm": 0.5346189141273499,
327
+ "learning_rate": 1e-06,
328
+ "loss": 0.6658,
329
+ "num_tokens": 39007950.0,
330
+ "step": 40
331
+ },
332
+ {
333
+ "epoch": 0.26537216828478966,
334
+ "grad_norm": 0.536400556564331,
335
+ "learning_rate": 1e-06,
336
+ "loss": 0.673,
337
+ "num_tokens": 40004404.0,
338
+ "step": 41
339
+ },
340
+ {
341
+ "epoch": 0.27184466019417475,
342
+ "grad_norm": 0.5293827652931213,
343
+ "learning_rate": 1e-06,
344
+ "loss": 0.6625,
345
+ "num_tokens": 40979367.0,
346
+ "step": 42
347
+ },
348
+ {
349
+ "epoch": 0.2783171521035599,
350
+ "grad_norm": 0.5181026458740234,
351
+ "learning_rate": 1e-06,
352
+ "loss": 0.6512,
353
+ "num_tokens": 41954716.0,
354
+ "step": 43
355
+ },
356
+ {
357
+ "epoch": 0.284789644012945,
358
+ "grad_norm": 0.4992174506187439,
359
+ "learning_rate": 1e-06,
360
+ "loss": 0.6621,
361
+ "num_tokens": 42927994.0,
362
+ "step": 44
363
+ },
364
+ {
365
+ "epoch": 0.2912621359223301,
366
+ "grad_norm": 0.47395315766334534,
367
+ "learning_rate": 1e-06,
368
+ "loss": 0.6478,
369
+ "num_tokens": 43870458.0,
370
+ "step": 45
371
+ },
372
+ {
373
+ "epoch": 0.2977346278317152,
374
+ "grad_norm": 0.41670018434524536,
375
+ "learning_rate": 1e-06,
376
+ "loss": 0.6411,
377
+ "num_tokens": 44845205.0,
378
+ "step": 46
379
+ },
380
+ {
381
+ "epoch": 0.3042071197411003,
382
+ "grad_norm": 0.3809607923030853,
383
+ "learning_rate": 1e-06,
384
+ "loss": 0.6372,
385
+ "num_tokens": 45817785.0,
386
+ "step": 47
387
+ },
388
+ {
389
+ "epoch": 0.3106796116504854,
390
+ "grad_norm": 0.36747702956199646,
391
+ "learning_rate": 1e-06,
392
+ "loss": 0.6403,
393
+ "num_tokens": 46768552.0,
394
+ "step": 48
395
+ },
396
+ {
397
+ "epoch": 0.31715210355987056,
398
+ "grad_norm": 0.34375864267349243,
399
+ "learning_rate": 1e-06,
400
+ "loss": 0.6193,
401
+ "num_tokens": 47720390.0,
402
+ "step": 49
403
+ },
404
+ {
405
+ "epoch": 0.32362459546925565,
406
+ "grad_norm": 0.3278081715106964,
407
+ "learning_rate": 1e-06,
408
+ "loss": 0.6277,
409
+ "num_tokens": 48659119.0,
410
+ "step": 50
411
+ },
412
+ {
413
+ "epoch": 0.3300970873786408,
414
+ "grad_norm": 0.3173641860485077,
415
+ "learning_rate": 1e-06,
416
+ "loss": 0.6262,
417
+ "num_tokens": 49635896.0,
418
+ "step": 51
419
+ },
420
+ {
421
+ "epoch": 0.3365695792880259,
422
+ "grad_norm": 0.3114546239376068,
423
+ "learning_rate": 1e-06,
424
+ "loss": 0.6229,
425
+ "num_tokens": 50617481.0,
426
+ "step": 52
427
+ },
428
+ {
429
+ "epoch": 0.343042071197411,
430
+ "grad_norm": 0.31268301606178284,
431
+ "learning_rate": 1e-06,
432
+ "loss": 0.6187,
433
+ "num_tokens": 51580322.0,
434
+ "step": 53
435
+ },
436
+ {
437
+ "epoch": 0.34951456310679613,
438
+ "grad_norm": 0.29753509163856506,
439
+ "learning_rate": 1e-06,
440
+ "loss": 0.6176,
441
+ "num_tokens": 52544983.0,
442
+ "step": 54
443
+ },
444
+ {
445
+ "epoch": 0.3559870550161812,
446
+ "grad_norm": 0.2916501760482788,
447
+ "learning_rate": 1e-06,
448
+ "loss": 0.6086,
449
+ "num_tokens": 53538431.0,
450
+ "step": 55
451
+ },
452
+ {
453
+ "epoch": 0.36245954692556637,
454
+ "grad_norm": 0.2825452983379364,
455
+ "learning_rate": 1e-06,
456
+ "loss": 0.6077,
457
+ "num_tokens": 54497097.0,
458
+ "step": 56
459
+ },
460
+ {
461
+ "epoch": 0.36893203883495146,
462
+ "grad_norm": 0.264753133058548,
463
+ "learning_rate": 1e-06,
464
+ "loss": 0.5966,
465
+ "num_tokens": 55515695.0,
466
+ "step": 57
467
+ },
468
+ {
469
+ "epoch": 0.37540453074433655,
470
+ "grad_norm": 0.25653982162475586,
471
+ "learning_rate": 1e-06,
472
+ "loss": 0.5997,
473
+ "num_tokens": 56471100.0,
474
+ "step": 58
475
+ },
476
+ {
477
+ "epoch": 0.3818770226537217,
478
+ "grad_norm": 0.2426663488149643,
479
+ "learning_rate": 1e-06,
480
+ "loss": 0.603,
481
+ "num_tokens": 57460948.0,
482
+ "step": 59
483
+ },
484
+ {
485
+ "epoch": 0.3883495145631068,
486
+ "grad_norm": 0.22974367439746857,
487
+ "learning_rate": 1e-06,
488
+ "loss": 0.5993,
489
+ "num_tokens": 58450663.0,
490
+ "step": 60
491
+ },
492
+ {
493
+ "epoch": 0.3948220064724919,
494
+ "grad_norm": 0.21884475648403168,
495
+ "learning_rate": 1e-06,
496
+ "loss": 0.5884,
497
+ "num_tokens": 59407638.0,
498
+ "step": 61
499
+ },
500
+ {
501
+ "epoch": 0.40129449838187703,
502
+ "grad_norm": 0.21299949288368225,
503
+ "learning_rate": 1e-06,
504
+ "loss": 0.5862,
505
+ "num_tokens": 60358388.0,
506
+ "step": 62
507
+ },
508
+ {
509
+ "epoch": 0.4077669902912621,
510
+ "grad_norm": 0.21099086105823517,
511
+ "learning_rate": 1e-06,
512
+ "loss": 0.5787,
513
+ "num_tokens": 61306137.0,
514
+ "step": 63
515
+ },
516
+ {
517
+ "epoch": 0.41423948220064727,
518
+ "grad_norm": 0.20446249842643738,
519
+ "learning_rate": 1e-06,
520
+ "loss": 0.5888,
521
+ "num_tokens": 62298616.0,
522
+ "step": 64
523
+ },
524
+ {
525
+ "epoch": 0.42071197411003236,
526
+ "grad_norm": 0.20021338760852814,
527
+ "learning_rate": 1e-06,
528
+ "loss": 0.5759,
529
+ "num_tokens": 63242620.0,
530
+ "step": 65
531
+ },
532
+ {
533
+ "epoch": 0.42718446601941745,
534
+ "grad_norm": 0.20259256660938263,
535
+ "learning_rate": 1e-06,
536
+ "loss": 0.5939,
537
+ "num_tokens": 64231555.0,
538
+ "step": 66
539
+ },
540
+ {
541
+ "epoch": 0.4336569579288026,
542
+ "grad_norm": 0.19663812220096588,
543
+ "learning_rate": 1e-06,
544
+ "loss": 0.5788,
545
+ "num_tokens": 65190302.0,
546
+ "step": 67
547
+ },
548
+ {
549
+ "epoch": 0.4401294498381877,
550
+ "grad_norm": 0.1927073448896408,
551
+ "learning_rate": 1e-06,
552
+ "loss": 0.5755,
553
+ "num_tokens": 66175194.0,
554
+ "step": 68
555
+ },
556
+ {
557
+ "epoch": 0.44660194174757284,
558
+ "grad_norm": 0.18935494124889374,
559
+ "learning_rate": 1e-06,
560
+ "loss": 0.573,
561
+ "num_tokens": 67146659.0,
562
+ "step": 69
563
+ },
564
+ {
565
+ "epoch": 0.45307443365695793,
566
+ "grad_norm": 0.18019847571849823,
567
+ "learning_rate": 1e-06,
568
+ "loss": 0.5539,
569
+ "num_tokens": 68115421.0,
570
+ "step": 70
571
+ },
572
+ {
573
+ "epoch": 0.459546925566343,
574
+ "grad_norm": 0.17649590969085693,
575
+ "learning_rate": 1e-06,
576
+ "loss": 0.5506,
577
+ "num_tokens": 69078962.0,
578
+ "step": 71
579
+ },
580
+ {
581
+ "epoch": 0.46601941747572817,
582
+ "grad_norm": 0.16998673975467682,
583
+ "learning_rate": 1e-06,
584
+ "loss": 0.573,
585
+ "num_tokens": 70056552.0,
586
+ "step": 72
587
+ },
588
+ {
589
+ "epoch": 0.47249190938511326,
590
+ "grad_norm": 0.16949011385440826,
591
+ "learning_rate": 1e-06,
592
+ "loss": 0.5674,
593
+ "num_tokens": 71056405.0,
594
+ "step": 73
595
+ },
596
+ {
597
+ "epoch": 0.47896440129449835,
598
+ "grad_norm": 0.1706332266330719,
599
+ "learning_rate": 1e-06,
600
+ "loss": 0.5755,
601
+ "num_tokens": 72011963.0,
602
+ "step": 74
603
+ },
604
+ {
605
+ "epoch": 0.4854368932038835,
606
+ "grad_norm": 0.16622762382030487,
607
+ "learning_rate": 1e-06,
608
+ "loss": 0.563,
609
+ "num_tokens": 72974088.0,
610
+ "step": 75
611
+ },
612
+ {
613
+ "epoch": 0.4919093851132686,
614
+ "grad_norm": 0.16113753616809845,
615
+ "learning_rate": 1e-06,
616
+ "loss": 0.5545,
617
+ "num_tokens": 73963066.0,
618
+ "step": 76
619
+ },
620
+ {
621
+ "epoch": 0.49838187702265374,
622
+ "grad_norm": 0.15716572105884552,
623
+ "learning_rate": 1e-06,
624
+ "loss": 0.5465,
625
+ "num_tokens": 74927507.0,
626
+ "step": 77
627
+ },
628
+ {
629
+ "epoch": 0.5048543689320388,
630
+ "grad_norm": 0.15952879190444946,
631
+ "learning_rate": 1e-06,
632
+ "loss": 0.5513,
633
+ "num_tokens": 75913302.0,
634
+ "step": 78
635
+ },
636
+ {
637
+ "epoch": 0.511326860841424,
638
+ "grad_norm": 0.1577850878238678,
639
+ "learning_rate": 1e-06,
640
+ "loss": 0.559,
641
+ "num_tokens": 76871637.0,
642
+ "step": 79
643
+ },
644
+ {
645
+ "epoch": 0.517799352750809,
646
+ "grad_norm": 0.1505461186170578,
647
+ "learning_rate": 1e-06,
648
+ "loss": 0.5509,
649
+ "num_tokens": 77844738.0,
650
+ "step": 80
651
+ },
652
+ {
653
+ "epoch": 0.5242718446601942,
654
+ "grad_norm": 0.1514582335948944,
655
+ "learning_rate": 1e-06,
656
+ "loss": 0.5453,
657
+ "num_tokens": 78789735.0,
658
+ "step": 81
659
+ },
660
+ {
661
+ "epoch": 0.5307443365695793,
662
+ "grad_norm": 0.15150229632854462,
663
+ "learning_rate": 1e-06,
664
+ "loss": 0.5417,
665
+ "num_tokens": 79740057.0,
666
+ "step": 82
667
+ },
668
+ {
669
+ "epoch": 0.5372168284789643,
670
+ "grad_norm": 0.14974889159202576,
671
+ "learning_rate": 1e-06,
672
+ "loss": 0.5498,
673
+ "num_tokens": 80691344.0,
674
+ "step": 83
675
+ },
676
+ {
677
+ "epoch": 0.5436893203883495,
678
+ "grad_norm": 0.14659319818019867,
679
+ "learning_rate": 1e-06,
680
+ "loss": 0.5456,
681
+ "num_tokens": 81695065.0,
682
+ "step": 84
683
+ },
684
+ {
685
+ "epoch": 0.5501618122977346,
686
+ "grad_norm": 0.14839723706245422,
687
+ "learning_rate": 1e-06,
688
+ "loss": 0.5446,
689
+ "num_tokens": 82670591.0,
690
+ "step": 85
691
+ },
692
+ {
693
+ "epoch": 0.5566343042071198,
694
+ "grad_norm": 0.1449030339717865,
695
+ "learning_rate": 1e-06,
696
+ "loss": 0.5434,
697
+ "num_tokens": 83619525.0,
698
+ "step": 86
699
+ },
700
+ {
701
+ "epoch": 0.5631067961165048,
702
+ "grad_norm": 0.14636728167533875,
703
+ "learning_rate": 1e-06,
704
+ "loss": 0.5427,
705
+ "num_tokens": 84583829.0,
706
+ "step": 87
707
+ },
708
+ {
709
+ "epoch": 0.56957928802589,
710
+ "grad_norm": 0.14410880208015442,
711
+ "learning_rate": 1e-06,
712
+ "loss": 0.54,
713
+ "num_tokens": 85554557.0,
714
+ "step": 88
715
+ },
716
+ {
717
+ "epoch": 0.5760517799352751,
718
+ "grad_norm": 0.14123332500457764,
719
+ "learning_rate": 1e-06,
720
+ "loss": 0.5412,
721
+ "num_tokens": 86534800.0,
722
+ "step": 89
723
+ },
724
+ {
725
+ "epoch": 0.5825242718446602,
726
+ "grad_norm": 0.14631222188472748,
727
+ "learning_rate": 1e-06,
728
+ "loss": 0.5451,
729
+ "num_tokens": 87479322.0,
730
+ "step": 90
731
+ },
732
+ {
733
+ "epoch": 0.5889967637540453,
734
+ "grad_norm": 0.141328364610672,
735
+ "learning_rate": 1e-06,
736
+ "loss": 0.543,
737
+ "num_tokens": 88440604.0,
738
+ "step": 91
739
+ },
740
+ {
741
+ "epoch": 0.5954692556634305,
742
+ "grad_norm": 0.14410269260406494,
743
+ "learning_rate": 1e-06,
744
+ "loss": 0.5389,
745
+ "num_tokens": 89440902.0,
746
+ "step": 92
747
+ },
748
+ {
749
+ "epoch": 0.6019417475728155,
750
+ "grad_norm": 0.14109647274017334,
751
+ "learning_rate": 1e-06,
752
+ "loss": 0.5312,
753
+ "num_tokens": 90420292.0,
754
+ "step": 93
755
+ },
756
+ {
757
+ "epoch": 0.6084142394822006,
758
+ "grad_norm": 0.14093320071697235,
759
+ "learning_rate": 1e-06,
760
+ "loss": 0.5432,
761
+ "num_tokens": 91365927.0,
762
+ "step": 94
763
+ },
764
+ {
765
+ "epoch": 0.6148867313915858,
766
+ "grad_norm": 0.1392253190279007,
767
+ "learning_rate": 1e-06,
768
+ "loss": 0.5342,
769
+ "num_tokens": 92339481.0,
770
+ "step": 95
771
+ },
772
+ {
773
+ "epoch": 0.6213592233009708,
774
+ "grad_norm": 0.14118197560310364,
775
+ "learning_rate": 1e-06,
776
+ "loss": 0.5291,
777
+ "num_tokens": 93277920.0,
778
+ "step": 96
779
+ },
780
+ {
781
+ "epoch": 0.627831715210356,
782
+ "grad_norm": 0.14340411126613617,
783
+ "learning_rate": 1e-06,
784
+ "loss": 0.527,
785
+ "num_tokens": 94233058.0,
786
+ "step": 97
787
+ },
788
+ {
789
+ "epoch": 0.6343042071197411,
790
+ "grad_norm": 0.13777285814285278,
791
+ "learning_rate": 1e-06,
792
+ "loss": 0.5407,
793
+ "num_tokens": 95195865.0,
794
+ "step": 98
795
+ },
796
+ {
797
+ "epoch": 0.6407766990291263,
798
+ "grad_norm": 0.14087024331092834,
799
+ "learning_rate": 1e-06,
800
+ "loss": 0.5219,
801
+ "num_tokens": 96165533.0,
802
+ "step": 99
803
+ },
804
+ {
805
+ "epoch": 0.6472491909385113,
806
+ "grad_norm": 0.13480006158351898,
807
+ "learning_rate": 1e-06,
808
+ "loss": 0.5229,
809
+ "num_tokens": 97128908.0,
810
+ "step": 100
811
+ },
812
+ {
813
+ "epoch": 0.6537216828478964,
814
+ "grad_norm": 0.13443224132061005,
815
+ "learning_rate": 1e-06,
816
+ "loss": 0.5235,
817
+ "num_tokens": 98137812.0,
818
+ "step": 101
819
+ },
820
+ {
821
+ "epoch": 0.6601941747572816,
822
+ "grad_norm": 0.13486695289611816,
823
+ "learning_rate": 1e-06,
824
+ "loss": 0.521,
825
+ "num_tokens": 99133445.0,
826
+ "step": 102
827
+ },
828
+ {
829
+ "epoch": 0.6666666666666666,
830
+ "grad_norm": 0.13491222262382507,
831
+ "learning_rate": 1e-06,
832
+ "loss": 0.5118,
833
+ "num_tokens": 100106034.0,
834
+ "step": 103
835
+ },
836
+ {
837
+ "epoch": 0.6731391585760518,
838
+ "grad_norm": 0.13977712392807007,
839
+ "learning_rate": 1e-06,
840
+ "loss": 0.5394,
841
+ "num_tokens": 101060830.0,
842
+ "step": 104
843
+ },
844
+ {
845
+ "epoch": 0.6796116504854369,
846
+ "grad_norm": 0.1345338374376297,
847
+ "learning_rate": 1e-06,
848
+ "loss": 0.5226,
849
+ "num_tokens": 102032708.0,
850
+ "step": 105
851
+ },
852
+ {
853
+ "epoch": 0.686084142394822,
854
+ "grad_norm": 0.13142578303813934,
855
+ "learning_rate": 1e-06,
856
+ "loss": 0.5061,
857
+ "num_tokens": 103020888.0,
858
+ "step": 106
859
+ },
860
+ {
861
+ "epoch": 0.6925566343042071,
862
+ "grad_norm": 0.14058642089366913,
863
+ "learning_rate": 1e-06,
864
+ "loss": 0.5315,
865
+ "num_tokens": 103986883.0,
866
+ "step": 107
867
+ },
868
+ {
869
+ "epoch": 0.6990291262135923,
870
+ "grad_norm": 0.13689804077148438,
871
+ "learning_rate": 1e-06,
872
+ "loss": 0.5275,
873
+ "num_tokens": 104978514.0,
874
+ "step": 108
875
+ },
876
+ {
877
+ "epoch": 0.7055016181229773,
878
+ "grad_norm": 0.1382923722267151,
879
+ "learning_rate": 1e-06,
880
+ "loss": 0.5159,
881
+ "num_tokens": 105938542.0,
882
+ "step": 109
883
+ },
884
+ {
885
+ "epoch": 0.7119741100323624,
886
+ "grad_norm": 0.13341419398784637,
887
+ "learning_rate": 1e-06,
888
+ "loss": 0.5175,
889
+ "num_tokens": 106909993.0,
890
+ "step": 110
891
+ },
892
+ {
893
+ "epoch": 0.7184466019417476,
894
+ "grad_norm": 0.13735052943229675,
895
+ "learning_rate": 1e-06,
896
+ "loss": 0.5204,
897
+ "num_tokens": 107872313.0,
898
+ "step": 111
899
+ },
900
+ {
901
+ "epoch": 0.7249190938511327,
902
+ "grad_norm": 0.14685890078544617,
903
+ "learning_rate": 1e-06,
904
+ "loss": 0.5154,
905
+ "num_tokens": 108876000.0,
906
+ "step": 112
907
+ },
908
+ {
909
+ "epoch": 0.7313915857605178,
910
+ "grad_norm": 0.13520538806915283,
911
+ "learning_rate": 1e-06,
912
+ "loss": 0.5163,
913
+ "num_tokens": 109857790.0,
914
+ "step": 113
915
+ },
916
+ {
917
+ "epoch": 0.7378640776699029,
918
+ "grad_norm": 0.1330995261669159,
919
+ "learning_rate": 1e-06,
920
+ "loss": 0.5155,
921
+ "num_tokens": 110826497.0,
922
+ "step": 114
923
+ },
924
+ {
925
+ "epoch": 0.7443365695792881,
926
+ "grad_norm": 0.134913370013237,
927
+ "learning_rate": 1e-06,
928
+ "loss": 0.5142,
929
+ "num_tokens": 111775780.0,
930
+ "step": 115
931
+ },
932
+ {
933
+ "epoch": 0.7508090614886731,
934
+ "grad_norm": 0.13504451513290405,
935
+ "learning_rate": 1e-06,
936
+ "loss": 0.5161,
937
+ "num_tokens": 112755203.0,
938
+ "step": 116
939
+ },
940
+ {
941
+ "epoch": 0.7572815533980582,
942
+ "grad_norm": 0.13805583119392395,
943
+ "learning_rate": 1e-06,
944
+ "loss": 0.5166,
945
+ "num_tokens": 113720983.0,
946
+ "step": 117
947
+ },
948
+ {
949
+ "epoch": 0.7637540453074434,
950
+ "grad_norm": 0.137771874666214,
951
+ "learning_rate": 1e-06,
952
+ "loss": 0.5057,
953
+ "num_tokens": 114690500.0,
954
+ "step": 118
955
+ },
956
+ {
957
+ "epoch": 0.7702265372168284,
958
+ "grad_norm": 0.13603690266609192,
959
+ "learning_rate": 1e-06,
960
+ "loss": 0.5059,
961
+ "num_tokens": 115663752.0,
962
+ "step": 119
963
+ },
964
+ {
965
+ "epoch": 0.7766990291262136,
966
+ "grad_norm": 0.1295483261346817,
967
+ "learning_rate": 1e-06,
968
+ "loss": 0.5031,
969
+ "num_tokens": 116656118.0,
970
+ "step": 120
971
+ },
972
+ {
973
+ "epoch": 0.7831715210355987,
974
+ "grad_norm": 0.13289949297904968,
975
+ "learning_rate": 1e-06,
976
+ "loss": 0.5127,
977
+ "num_tokens": 117647006.0,
978
+ "step": 121
979
+ },
980
+ {
981
+ "epoch": 0.7896440129449838,
982
+ "grad_norm": 0.13150979578495026,
983
+ "learning_rate": 1e-06,
984
+ "loss": 0.5023,
985
+ "num_tokens": 118637898.0,
986
+ "step": 122
987
+ },
988
+ {
989
+ "epoch": 0.7961165048543689,
990
+ "grad_norm": 0.13312214612960815,
991
+ "learning_rate": 1e-06,
992
+ "loss": 0.5143,
993
+ "num_tokens": 119611539.0,
994
+ "step": 123
995
+ },
996
+ {
997
+ "epoch": 0.8025889967637541,
998
+ "grad_norm": 0.13679178059101105,
999
+ "learning_rate": 1e-06,
1000
+ "loss": 0.5149,
1001
+ "num_tokens": 120576485.0,
1002
+ "step": 124
1003
+ },
1004
+ {
1005
+ "epoch": 0.8090614886731392,
1006
+ "grad_norm": 0.13002805411815643,
1007
+ "learning_rate": 1e-06,
1008
+ "loss": 0.5143,
1009
+ "num_tokens": 121567862.0,
1010
+ "step": 125
1011
+ },
1012
+ {
1013
+ "epoch": 0.8155339805825242,
1014
+ "grad_norm": 0.13436037302017212,
1015
+ "learning_rate": 1e-06,
1016
+ "loss": 0.5028,
1017
+ "num_tokens": 122530893.0,
1018
+ "step": 126
1019
+ },
1020
+ {
1021
+ "epoch": 0.8220064724919094,
1022
+ "grad_norm": 0.13467973470687866,
1023
+ "learning_rate": 1e-06,
1024
+ "loss": 0.5069,
1025
+ "num_tokens": 123480227.0,
1026
+ "step": 127
1027
+ },
1028
+ {
1029
+ "epoch": 0.8284789644012945,
1030
+ "grad_norm": 0.13601741194725037,
1031
+ "learning_rate": 1e-06,
1032
+ "loss": 0.5194,
1033
+ "num_tokens": 124462070.0,
1034
+ "step": 128
1035
+ },
1036
+ {
1037
+ "epoch": 0.8349514563106796,
1038
+ "grad_norm": 0.1381833255290985,
1039
+ "learning_rate": 1e-06,
1040
+ "loss": 0.5142,
1041
+ "num_tokens": 125440428.0,
1042
+ "step": 129
1043
+ },
1044
+ {
1045
+ "epoch": 0.8414239482200647,
1046
+ "grad_norm": 0.1346794068813324,
1047
+ "learning_rate": 1e-06,
1048
+ "loss": 0.5115,
1049
+ "num_tokens": 126416310.0,
1050
+ "step": 130
1051
+ },
1052
+ {
1053
+ "epoch": 0.8478964401294499,
1054
+ "grad_norm": 0.13682374358177185,
1055
+ "learning_rate": 1e-06,
1056
+ "loss": 0.512,
1057
+ "num_tokens": 127406436.0,
1058
+ "step": 131
1059
+ },
1060
+ {
1061
+ "epoch": 0.8543689320388349,
1062
+ "grad_norm": 0.13239574432373047,
1063
+ "learning_rate": 1e-06,
1064
+ "loss": 0.5068,
1065
+ "num_tokens": 128372197.0,
1066
+ "step": 132
1067
+ },
1068
+ {
1069
+ "epoch": 0.86084142394822,
1070
+ "grad_norm": 0.13347913324832916,
1071
+ "learning_rate": 1e-06,
1072
+ "loss": 0.5062,
1073
+ "num_tokens": 129371264.0,
1074
+ "step": 133
1075
+ },
1076
+ {
1077
+ "epoch": 0.8673139158576052,
1078
+ "grad_norm": 0.1342724710702896,
1079
+ "learning_rate": 1e-06,
1080
+ "loss": 0.4917,
1081
+ "num_tokens": 130306080.0,
1082
+ "step": 134
1083
+ },
1084
+ {
1085
+ "epoch": 0.8737864077669902,
1086
+ "grad_norm": 0.13179434835910797,
1087
+ "learning_rate": 1e-06,
1088
+ "loss": 0.5107,
1089
+ "num_tokens": 131261937.0,
1090
+ "step": 135
1091
+ },
1092
+ {
1093
+ "epoch": 0.8802588996763754,
1094
+ "grad_norm": 0.13348270952701569,
1095
+ "learning_rate": 1e-06,
1096
+ "loss": 0.5044,
1097
+ "num_tokens": 132223182.0,
1098
+ "step": 136
1099
+ },
1100
+ {
1101
+ "epoch": 0.8867313915857605,
1102
+ "grad_norm": 0.22263851761817932,
1103
+ "learning_rate": 1e-06,
1104
+ "loss": 0.5082,
1105
+ "num_tokens": 133211562.0,
1106
+ "step": 137
1107
+ },
1108
+ {
1109
+ "epoch": 0.8932038834951457,
1110
+ "grad_norm": 0.13348044455051422,
1111
+ "learning_rate": 1e-06,
1112
+ "loss": 0.5043,
1113
+ "num_tokens": 134208133.0,
1114
+ "step": 138
1115
+ },
1116
+ {
1117
+ "epoch": 0.8996763754045307,
1118
+ "grad_norm": 0.13663546741008759,
1119
+ "learning_rate": 1e-06,
1120
+ "loss": 0.5083,
1121
+ "num_tokens": 135194807.0,
1122
+ "step": 139
1123
+ },
1124
+ {
1125
+ "epoch": 0.9061488673139159,
1126
+ "grad_norm": 0.1293790638446808,
1127
+ "learning_rate": 1e-06,
1128
+ "loss": 0.5046,
1129
+ "num_tokens": 136175941.0,
1130
+ "step": 140
1131
+ },
1132
+ {
1133
+ "epoch": 0.912621359223301,
1134
+ "grad_norm": 0.13469921052455902,
1135
+ "learning_rate": 1e-06,
1136
+ "loss": 0.5183,
1137
+ "num_tokens": 137117638.0,
1138
+ "step": 141
1139
+ },
1140
+ {
1141
+ "epoch": 0.919093851132686,
1142
+ "grad_norm": 0.1336260437965393,
1143
+ "learning_rate": 1e-06,
1144
+ "loss": 0.4941,
1145
+ "num_tokens": 138102412.0,
1146
+ "step": 142
1147
+ },
1148
+ {
1149
+ "epoch": 0.9255663430420712,
1150
+ "grad_norm": 0.1314634084701538,
1151
+ "learning_rate": 1e-06,
1152
+ "loss": 0.497,
1153
+ "num_tokens": 139056304.0,
1154
+ "step": 143
1155
+ },
1156
+ {
1157
+ "epoch": 0.9320388349514563,
1158
+ "grad_norm": 0.13598428666591644,
1159
+ "learning_rate": 1e-06,
1160
+ "loss": 0.5026,
1161
+ "num_tokens": 140002988.0,
1162
+ "step": 144
1163
+ },
1164
+ {
1165
+ "epoch": 0.9385113268608414,
1166
+ "grad_norm": 0.13465505838394165,
1167
+ "learning_rate": 1e-06,
1168
+ "loss": 0.5051,
1169
+ "num_tokens": 140939323.0,
1170
+ "step": 145
1171
+ },
1172
+ {
1173
+ "epoch": 0.9449838187702265,
1174
+ "grad_norm": 0.13320808112621307,
1175
+ "learning_rate": 1e-06,
1176
+ "loss": 0.5007,
1177
+ "num_tokens": 141898246.0,
1178
+ "step": 146
1179
+ },
1180
+ {
1181
+ "epoch": 0.9514563106796117,
1182
+ "grad_norm": 0.13584767282009125,
1183
+ "learning_rate": 1e-06,
1184
+ "loss": 0.5074,
1185
+ "num_tokens": 142861118.0,
1186
+ "step": 147
1187
+ },
1188
+ {
1189
+ "epoch": 0.9579288025889967,
1190
+ "grad_norm": 0.13394641876220703,
1191
+ "learning_rate": 1e-06,
1192
+ "loss": 0.5002,
1193
+ "num_tokens": 143845628.0,
1194
+ "step": 148
1195
+ },
1196
+ {
1197
+ "epoch": 0.9644012944983819,
1198
+ "grad_norm": 0.13663163781166077,
1199
+ "learning_rate": 1e-06,
1200
+ "loss": 0.5129,
1201
+ "num_tokens": 144816454.0,
1202
+ "step": 149
1203
+ },
1204
+ {
1205
+ "epoch": 0.970873786407767,
1206
+ "grad_norm": 0.12985925376415253,
1207
+ "learning_rate": 1e-06,
1208
+ "loss": 0.5006,
1209
+ "num_tokens": 145791528.0,
1210
+ "step": 150
1211
+ },
1212
+ {
1213
+ "epoch": 0.9773462783171522,
1214
+ "grad_norm": 0.1344696283340454,
1215
+ "learning_rate": 1e-06,
1216
+ "loss": 0.5136,
1217
+ "num_tokens": 146776225.0,
1218
+ "step": 151
1219
+ },
1220
+ {
1221
+ "epoch": 0.9838187702265372,
1222
+ "grad_norm": 0.1393062323331833,
1223
+ "learning_rate": 1e-06,
1224
+ "loss": 0.4963,
1225
+ "num_tokens": 147716038.0,
1226
+ "step": 152
1227
+ },
1228
+ {
1229
+ "epoch": 0.9902912621359223,
1230
+ "grad_norm": 0.13568711280822754,
1231
+ "learning_rate": 1e-06,
1232
+ "loss": 0.4935,
1233
+ "num_tokens": 148641036.0,
1234
+ "step": 153
1235
+ },
1236
+ {
1237
+ "epoch": 0.9967637540453075,
1238
+ "grad_norm": 0.12916181981563568,
1239
+ "learning_rate": 1e-06,
1240
+ "loss": 0.4964,
1241
+ "num_tokens": 149611850.0,
1242
+ "step": 154
1243
+ },
1244
+ {
1245
+ "epoch": 0.9967637540453075,
1246
+ "step": 154,
1247
+ "total_flos": 6.182777521931551e+18,
1248
+ "train_loss": 0.6460230693027571,
1249
+ "train_runtime": 6408.7951,
1250
+ "train_samples_per_second": 10.787,
1251
+ "train_steps_per_second": 0.024
1252
+ }
1253
+ ],
1254
+ "logging_steps": 1,
1255
+ "max_steps": 154,
1256
+ "num_input_tokens_seen": 0,
1257
+ "num_train_epochs": 1,
1258
+ "save_steps": 8,
1259
+ "stateful_callbacks": {
1260
+ "TrainerControl": {
1261
+ "args": {
1262
+ "should_epoch_stop": false,
1263
+ "should_evaluate": false,
1264
+ "should_log": false,
1265
+ "should_save": true,
1266
+ "should_training_stop": true
1267
+ },
1268
+ "attributes": {}
1269
+ }
1270
+ },
1271
+ "total_flos": 6.182777521931551e+18,
1272
+ "train_batch_size": 16,
1273
+ "trial_name": null,
1274
+ "trial_params": null
1275
+ }