Hugging Face's logo

onuralp
/
Llama-2-13b-QLoRA-02

Text Generation
Transformers
llama
Generated from Trainer
Inference Endpoints
4-bit precision
bitsandbytes
Model card Files Community
onuralp commited on
Commit
df50375
•
1 Parent(s): e436853

Upload folder using huggingface_hub

Browse files
Files changed (15) hide show
  1. README.md +93 -0
  2. adapter_config.json +28 -0
  3. adapter_model.bin +3 -0
  4. checkpoint-699/README.md +219 -0
  5. checkpoint-699/adapter_config.json +28 -0
  6. checkpoint-699/adapter_model.bin +3 -0
  7. checkpoint-699/optimizer.pt +3 -0
  8. checkpoint-699/rng_state.pth +3 -0
  9. checkpoint-699/scheduler.pt +3 -0
  10. checkpoint-699/trainer_state.json +4485 -0
  11. checkpoint-699/training_args.bin +3 -0
  12. config.json +39 -0
  13. special_tokens_map.json +24 -0
  14. tokenizer.model +3 -0
  15. tokenizer_config.json +44 -0
README.md ADDED
@@ -0,0 +1,93 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: meta-llama/Llama-2-13b-hf
3
+ tags:
4
+ - generated_from_trainer
5
+ model-index:
6
+ - name: qlora-out
7
+ results: []
8
+ ---
9
+
10
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
11
+ should probably proofread and complete it, then remove this comment. -->
12
+
13
+ [<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
14
+ # qlora-out
15
+
16
+ This model is a fine-tuned version of [meta-llama/Llama-2-13b-hf](https://huggingface.co/meta-llama/Llama-2-13b-hf) on the None dataset.
17
+ It achieves the following results on the evaluation set:
18
+ - Loss: 0.5407
19
+
20
+ ## Model description
21
+
22
+ More information needed
23
+
24
+ ## Intended uses & limitations
25
+
26
+ More information needed
27
+
28
+ ## Training and evaluation data
29
+
30
+ More information needed
31
+
32
+ ## Training procedure
33
+
34
+ ### Training hyperparameters
35
+
36
+ The following hyperparameters were used during training:
37
+ - learning_rate: 0.0004
38
+ - train_batch_size: 2
39
+ - eval_batch_size: 2
40
+ - seed: 42
41
+ - gradient_accumulation_steps: 4
42
+ - total_train_batch_size: 8
43
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
44
+ - lr_scheduler_type: cosine
45
+ - lr_scheduler_warmup_steps: 300
46
+ - num_epochs: 1
47
+
48
+ ### Training results
49
+
50
+ | Training Loss | Epoch | Step | Validation Loss |
51
+ |:-------------:|:-----:|:----:|:---------------:|
52
+ | 0.8973 | 0.03 | 20 | 0.7029 |
53
+ | 0.6828 | 0.06 | 40 | 0.6521 |
54
+ | 0.6521 | 0.09 | 60 | 0.6199 |
55
+ | 0.7857 | 0.11 | 80 | 0.6066 |
56
+ | 0.6208 | 0.14 | 100 | 0.6063 |
57
+ | 0.6805 | 0.17 | 120 | 0.5969 |
58
+ | 0.5928 | 0.2 | 140 | 0.5989 |
59
+ | 0.715 | 0.23 | 160 | 0.5844 |
60
+ | 0.5647 | 0.26 | 180 | 0.5979 |
61
+ | 0.6778 | 0.29 | 200 | 0.5889 |
62
+ | 0.5907 | 0.31 | 220 | 0.5772 |
63
+ | 0.5536 | 0.34 | 240 | 0.5917 |
64
+ | 0.7422 | 0.37 | 260 | 0.6781 |
65
+ | 0.6328 | 0.4 | 280 | 0.5785 |
66
+ | 0.5705 | 0.43 | 300 | 0.5720 |
67
+ | 0.6124 | 0.46 | 320 | 0.5753 |
68
+ | 0.4735 | 0.49 | 340 | 0.6203 |
69
+ | 0.4602 | 0.52 | 360 | 0.5772 |
70
+ | 0.8571 | 0.54 | 380 | 0.5750 |
71
+ | 0.5504 | 0.57 | 400 | 0.6040 |
72
+ | 0.6307 | 0.6 | 420 | 0.5796 |
73
+ | 0.4782 | 0.63 | 440 | 0.5639 |
74
+ | 0.4159 | 0.66 | 460 | 0.5689 |
75
+ | 0.6393 | 0.69 | 480 | 0.5661 |
76
+ | 0.8243 | 0.72 | 500 | 0.5698 |
77
+ | 0.4744 | 0.74 | 520 | 0.5536 |
78
+ | 0.4395 | 0.77 | 540 | 0.5536 |
79
+ | 0.543 | 0.8 | 560 | 0.5493 |
80
+ | 0.4451 | 0.83 | 580 | 0.5421 |
81
+ | 0.5384 | 0.86 | 600 | 0.5467 |
82
+ | 0.4438 | 0.89 | 620 | 0.5379 |
83
+ | 0.4168 | 0.92 | 640 | 0.5398 |
84
+ | 0.469 | 0.94 | 660 | 0.5402 |
85
+ | 0.6766 | 0.97 | 680 | 0.5407 |
86
+
87
+
88
+ ### Framework versions
89
+
90
+ - Transformers 4.34.1
91
+ - Pytorch 2.0.1+cu118
92
+ - Datasets 2.14.6
93
+ - Tokenizers 0.14.1
adapter_config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "meta-llama/Llama-2-13b-hf",
5
+ "bias": "none",
6
+ "fan_in_fan_out": null,
7
+ "inference_mode": true,
8
+ "init_lora_weights": true,
9
+ "layers_pattern": null,
10
+ "layers_to_transform": null,
11
+ "lora_alpha": 64,
12
+ "lora_dropout": 0.05,
13
+ "modules_to_save": null,
14
+ "peft_type": "LORA",
15
+ "r": 64,
16
+ "rank_pattern": {},
17
+ "revision": null,
18
+ "target_modules": [
19
+ "up_proj",
20
+ "gate_proj",
21
+ "o_proj",
22
+ "k_proj",
23
+ "q_proj",
24
+ "down_proj",
25
+ "v_proj"
26
+ ],
27
+ "task_type": "CAUSAL_LM"
28
+ }
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f792b2ba0e15ec006475a9e69f9332ca09a5b8c5c1f370aecf73fd09858fbe4b
3
+ size 1001592141
checkpoint-699/README.md ADDED
@@ -0,0 +1,219 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ base_model: meta-llama/Llama-2-13b-hf
4
+ ---
5
+
6
+ # Model Card for Model ID
7
+
8
+ <!-- Provide a quick summary of what the model is/does. -->
9
+
10
+
11
+
12
+ ## Model Details
13
+
14
+ ### Model Description
15
+
16
+ <!-- Provide a longer summary of what this model is. -->
17
+
18
+
19
+
20
+ - **Developed by:** [More Information Needed]
21
+ - **Shared by [optional]:** [More Information Needed]
22
+ - **Model type:** [More Information Needed]
23
+ - **Language(s) (NLP):** [More Information Needed]
24
+ - **License:** [More Information Needed]
25
+ - **Finetuned from model [optional]:** [More Information Needed]
26
+
27
+ ### Model Sources [optional]
28
+
29
+ <!-- Provide the basic links for the model. -->
30
+
31
+ - **Repository:** [More Information Needed]
32
+ - **Paper [optional]:** [More Information Needed]
33
+ - **Demo [optional]:** [More Information Needed]
34
+
35
+ ## Uses
36
+
37
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
38
+
39
+ ### Direct Use
40
+
41
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
42
+
43
+ [More Information Needed]
44
+
45
+ ### Downstream Use [optional]
46
+
47
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
48
+
49
+ [More Information Needed]
50
+
51
+ ### Out-of-Scope Use
52
+
53
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
54
+
55
+ [More Information Needed]
56
+
57
+ ## Bias, Risks, and Limitations
58
+
59
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
60
+
61
+ [More Information Needed]
62
+
63
+ ### Recommendations
64
+
65
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
66
+
67
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
68
+
69
+ ## How to Get Started with the Model
70
+
71
+ Use the code below to get started with the model.
72
+
73
+ [More Information Needed]
74
+
75
+ ## Training Details
76
+
77
+ ### Training Data
78
+
79
+ <!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
80
+
81
+ [More Information Needed]
82
+
83
+ ### Training Procedure
84
+
85
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
86
+
87
+ #### Preprocessing [optional]
88
+
89
+ [More Information Needed]
90
+
91
+
92
+ #### Training Hyperparameters
93
+
94
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
95
+
96
+ #### Speeds, Sizes, Times [optional]
97
+
98
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
99
+
100
+ [More Information Needed]
101
+
102
+ ## Evaluation
103
+
104
+ <!-- This section describes the evaluation protocols and provides the results. -->
105
+
106
+ ### Testing Data, Factors & Metrics
107
+
108
+ #### Testing Data
109
+
110
+ <!-- This should link to a Data Card if possible. -->
111
+
112
+ [More Information Needed]
113
+
114
+ #### Factors
115
+
116
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
117
+
118
+ [More Information Needed]
119
+
120
+ #### Metrics
121
+
122
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
123
+
124
+ [More Information Needed]
125
+
126
+ ### Results
127
+
128
+ [More Information Needed]
129
+
130
+ #### Summary
131
+
132
+
133
+
134
+ ## Model Examination [optional]
135
+
136
+ <!-- Relevant interpretability work for the model goes here -->
137
+
138
+ [More Information Needed]
139
+
140
+ ## Environmental Impact
141
+
142
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
143
+
144
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
145
+
146
+ - **Hardware Type:** [More Information Needed]
147
+ - **Hours used:** [More Information Needed]
148
+ - **Cloud Provider:** [More Information Needed]
149
+ - **Compute Region:** [More Information Needed]
150
+ - **Carbon Emitted:** [More Information Needed]
151
+
152
+ ## Technical Specifications [optional]
153
+
154
+ ### Model Architecture and Objective
155
+
156
+ [More Information Needed]
157
+
158
+ ### Compute Infrastructure
159
+
160
+ [More Information Needed]
161
+
162
+ #### Hardware
163
+
164
+ [More Information Needed]
165
+
166
+ #### Software
167
+
168
+ [More Information Needed]
169
+
170
+ ## Citation [optional]
171
+
172
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
173
+
174
+ **BibTeX:**
175
+
176
+ [More Information Needed]
177
+
178
+ **APA:**
179
+
180
+ [More Information Needed]
181
+
182
+ ## Glossary [optional]
183
+
184
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
185
+
186
+ [More Information Needed]
187
+
188
+ ## More Information [optional]
189
+
190
+ [More Information Needed]
191
+
192
+ ## Model Card Authors [optional]
193
+
194
+ [More Information Needed]
195
+
196
+ ## Model Card Contact
197
+
198
+ [More Information Needed]
199
+
200
+
201
+ ## Training procedure
202
+
203
+
204
+ The following `bitsandbytes` quantization config was used during training:
205
+ - quant_method: bitsandbytes
206
+ - load_in_8bit: False
207
+ - load_in_4bit: True
208
+ - llm_int8_threshold: 6.0
209
+ - llm_int8_skip_modules: None
210
+ - llm_int8_enable_fp32_cpu_offload: False
211
+ - llm_int8_has_fp16_weight: False
212
+ - bnb_4bit_quant_type: nf4
213
+ - bnb_4bit_use_double_quant: True
214
+ - bnb_4bit_compute_dtype: bfloat16
215
+
216
+ ### Framework versions
217
+
218
+
219
+ - PEFT 0.6.0.dev0
checkpoint-699/adapter_config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "meta-llama/Llama-2-13b-hf",
5
+ "bias": "none",
6
+ "fan_in_fan_out": null,
7
+ "inference_mode": true,
8
+ "init_lora_weights": true,
9
+ "layers_pattern": null,
10
+ "layers_to_transform": null,
11
+ "lora_alpha": 64,
12
+ "lora_dropout": 0.05,
13
+ "modules_to_save": null,
14
+ "peft_type": "LORA",
15
+ "r": 64,
16
+ "rank_pattern": {},
17
+ "revision": null,
18
+ "target_modules": [
19
+ "up_proj",
20
+ "gate_proj",
21
+ "o_proj",
22
+ "k_proj",
23
+ "q_proj",
24
+ "down_proj",
25
+ "v_proj"
26
+ ],
27
+ "task_type": "CAUSAL_LM"
28
+ }
checkpoint-699/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f792b2ba0e15ec006475a9e69f9332ca09a5b8c5c1f370aecf73fd09858fbe4b
3
+ size 1001592141
checkpoint-699/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7f5e0a06fae221f114e7583c4eb60e00a6d25fa0fa5aaef248d8ce17afa0d972
3
+ size 2003127101
checkpoint-699/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e31024f56d99e5c4fa16b2dbeaab3fb8070ae8483287937553c5bca752f37f71
3
+ size 14575
checkpoint-699/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:355d0ad21374ed60706a6227d183abaeeeec565080a61c3198ecbfd4ae947a25
3
+ size 627
checkpoint-699/trainer_state.json ADDED
@@ -0,0 +1,4485 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "eval_steps": 20,
6
+ "global_step": 699,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 1.3333333333333334e-06,
14
+ "loss": 3.6582,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 2.666666666666667e-06,
20
+ "loss": 2.2305,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.0,
25
+ "learning_rate": 4.000000000000001e-06,
26
+ "loss": 1.5487,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.01,
31
+ "learning_rate": 5.333333333333334e-06,
32
+ "loss": 3.6515,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.01,
37
+ "learning_rate": 6.666666666666667e-06,
38
+ "loss": 3.7426,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.01,
43
+ "learning_rate": 8.000000000000001e-06,
44
+ "loss": 3.4014,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.01,
49
+ "learning_rate": 9.333333333333334e-06,
50
+ "loss": 3.2163,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.01,
55
+ "learning_rate": 1.0666666666666667e-05,
56
+ "loss": 3.823,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.01,
61
+ "learning_rate": 1.2e-05,
62
+ "loss": 4.3112,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.01,
67
+ "learning_rate": 1.3333333333333333e-05,
68
+ "loss": 2.5267,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.02,
73
+ "learning_rate": 1.4666666666666668e-05,
74
+ "loss": 3.2481,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.02,
79
+ "learning_rate": 1.6000000000000003e-05,
80
+ "loss": 2.4184,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.02,
85
+ "learning_rate": 1.7333333333333336e-05,
86
+ "loss": 2.7559,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.02,
91
+ "learning_rate": 1.866666666666667e-05,
92
+ "loss": 1.4468,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.02,
97
+ "learning_rate": 2e-05,
98
+ "loss": 1.4686,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.02,
103
+ "learning_rate": 2.1333333333333335e-05,
104
+ "loss": 1.1665,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.02,
109
+ "learning_rate": 2.2666666666666668e-05,
110
+ "loss": 0.9044,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 0.03,
115
+ "learning_rate": 2.4e-05,
116
+ "loss": 0.9294,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 0.03,
121
+ "learning_rate": 2.5333333333333337e-05,
122
+ "loss": 0.8645,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 0.03,
127
+ "learning_rate": 2.6666666666666667e-05,
128
+ "loss": 0.8973,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 0.03,
133
+ "eval_loss": 0.7028617262840271,
134
+ "eval_runtime": 76.0257,
135
+ "eval_samples_per_second": 21.427,
136
+ "eval_steps_per_second": 10.72,
137
+ "step": 20
138
+ },
139
+ {
140
+ "epoch": 0.03,
141
+ "learning_rate": 2.8000000000000003e-05,
142
+ "loss": 0.8187,
143
+ "step": 21
144
+ },
145
+ {
146
+ "epoch": 0.03,
147
+ "learning_rate": 2.9333333333333336e-05,
148
+ "loss": 0.9868,
149
+ "step": 22
150
+ },
151
+ {
152
+ "epoch": 0.03,
153
+ "learning_rate": 3.066666666666667e-05,
154
+ "loss": 0.8922,
155
+ "step": 23
156
+ },
157
+ {
158
+ "epoch": 0.03,
159
+ "learning_rate": 3.2000000000000005e-05,
160
+ "loss": 0.8536,
161
+ "step": 24
162
+ },
163
+ {
164
+ "epoch": 0.04,
165
+ "learning_rate": 3.3333333333333335e-05,
166
+ "loss": 0.9366,
167
+ "step": 25
168
+ },
169
+ {
170
+ "epoch": 0.04,
171
+ "learning_rate": 3.466666666666667e-05,
172
+ "loss": 1.0178,
173
+ "step": 26
174
+ },
175
+ {
176
+ "epoch": 0.04,
177
+ "learning_rate": 3.6e-05,
178
+ "loss": 1.1587,
179
+ "step": 27
180
+ },
181
+ {
182
+ "epoch": 0.04,
183
+ "learning_rate": 3.733333333333334e-05,
184
+ "loss": 0.7203,
185
+ "step": 28
186
+ },
187
+ {
188
+ "epoch": 0.04,
189
+ "learning_rate": 3.866666666666667e-05,
190
+ "loss": 0.6773,
191
+ "step": 29
192
+ },
193
+ {
194
+ "epoch": 0.04,
195
+ "learning_rate": 4e-05,
196
+ "loss": 1.0454,
197
+ "step": 30
198
+ },
199
+ {
200
+ "epoch": 0.04,
201
+ "learning_rate": 4.133333333333333e-05,
202
+ "loss": 0.8861,
203
+ "step": 31
204
+ },
205
+ {
206
+ "epoch": 0.05,
207
+ "learning_rate": 4.266666666666667e-05,
208
+ "loss": 0.9398,
209
+ "step": 32
210
+ },
211
+ {
212
+ "epoch": 0.05,
213
+ "learning_rate": 4.4000000000000006e-05,
214
+ "loss": 0.9018,
215
+ "step": 33
216
+ },
217
+ {
218
+ "epoch": 0.05,
219
+ "learning_rate": 4.5333333333333335e-05,
220
+ "loss": 0.9051,
221
+ "step": 34
222
+ },
223
+ {
224
+ "epoch": 0.05,
225
+ "learning_rate": 4.666666666666667e-05,
226
+ "loss": 0.7858,
227
+ "step": 35
228
+ },
229
+ {
230
+ "epoch": 0.05,
231
+ "learning_rate": 4.8e-05,
232
+ "loss": 0.6311,
233
+ "step": 36
234
+ },
235
+ {
236
+ "epoch": 0.05,
237
+ "learning_rate": 4.933333333333334e-05,
238
+ "loss": 0.7227,
239
+ "step": 37
240
+ },
241
+ {
242
+ "epoch": 0.05,
243
+ "learning_rate": 5.0666666666666674e-05,
244
+ "loss": 0.8281,
245
+ "step": 38
246
+ },
247
+ {
248
+ "epoch": 0.06,
249
+ "learning_rate": 5.2000000000000004e-05,
250
+ "loss": 0.7919,
251
+ "step": 39
252
+ },
253
+ {
254
+ "epoch": 0.06,
255
+ "learning_rate": 5.333333333333333e-05,
256
+ "loss": 0.6828,
257
+ "step": 40
258
+ },
259
+ {
260
+ "epoch": 0.06,
261
+ "eval_loss": 0.6520687341690063,
262
+ "eval_runtime": 75.9058,
263
+ "eval_samples_per_second": 21.461,
264
+ "eval_steps_per_second": 10.737,
265
+ "step": 40
266
+ },
267
+ {
268
+ "epoch": 0.06,
269
+ "learning_rate": 5.466666666666666e-05,
270
+ "loss": 0.9088,
271
+ "step": 41
272
+ },
273
+ {
274
+ "epoch": 0.06,
275
+ "learning_rate": 5.6000000000000006e-05,
276
+ "loss": 0.6432,
277
+ "step": 42
278
+ },
279
+ {
280
+ "epoch": 0.06,
281
+ "learning_rate": 5.7333333333333336e-05,
282
+ "loss": 0.749,
283
+ "step": 43
284
+ },
285
+ {
286
+ "epoch": 0.06,
287
+ "learning_rate": 5.866666666666667e-05,
288
+ "loss": 0.7316,
289
+ "step": 44
290
+ },
291
+ {
292
+ "epoch": 0.06,
293
+ "learning_rate": 6e-05,
294
+ "loss": 0.6837,
295
+ "step": 45
296
+ },
297
+ {
298
+ "epoch": 0.07,
299
+ "learning_rate": 6.133333333333334e-05,
300
+ "loss": 0.717,
301
+ "step": 46
302
+ },
303
+ {
304
+ "epoch": 0.07,
305
+ "learning_rate": 6.266666666666667e-05,
306
+ "loss": 0.728,
307
+ "step": 47
308
+ },
309
+ {
310
+ "epoch": 0.07,
311
+ "learning_rate": 6.400000000000001e-05,
312
+ "loss": 0.8732,
313
+ "step": 48
314
+ },
315
+ {
316
+ "epoch": 0.07,
317
+ "learning_rate": 6.533333333333334e-05,
318
+ "loss": 0.6397,
319
+ "step": 49
320
+ },
321
+ {
322
+ "epoch": 0.07,
323
+ "learning_rate": 6.666666666666667e-05,
324
+ "loss": 0.654,
325
+ "step": 50
326
+ },
327
+ {
328
+ "epoch": 0.07,
329
+ "learning_rate": 6.800000000000001e-05,
330
+ "loss": 0.7031,
331
+ "step": 51
332
+ },
333
+ {
334
+ "epoch": 0.07,
335
+ "learning_rate": 6.933333333333334e-05,
336
+ "loss": 0.7092,
337
+ "step": 52
338
+ },
339
+ {
340
+ "epoch": 0.08,
341
+ "learning_rate": 7.066666666666667e-05,
342
+ "loss": 0.9424,
343
+ "step": 53
344
+ },
345
+ {
346
+ "epoch": 0.08,
347
+ "learning_rate": 7.2e-05,
348
+ "loss": 0.7815,
349
+ "step": 54
350
+ },
351
+ {
352
+ "epoch": 0.08,
353
+ "learning_rate": 7.333333333333333e-05,
354
+ "loss": 0.6843,
355
+ "step": 55
356
+ },
357
+ {
358
+ "epoch": 0.08,
359
+ "learning_rate": 7.466666666666667e-05,
360
+ "loss": 0.6269,
361
+ "step": 56
362
+ },
363
+ {
364
+ "epoch": 0.08,
365
+ "learning_rate": 7.6e-05,
366
+ "loss": 0.6285,
367
+ "step": 57
368
+ },
369
+ {
370
+ "epoch": 0.08,
371
+ "learning_rate": 7.733333333333333e-05,
372
+ "loss": 0.7555,
373
+ "step": 58
374
+ },
375
+ {
376
+ "epoch": 0.08,
377
+ "learning_rate": 7.866666666666666e-05,
378
+ "loss": 0.6724,
379
+ "step": 59
380
+ },
381
+ {
382
+ "epoch": 0.09,
383
+ "learning_rate": 8e-05,
384
+ "loss": 0.6521,
385
+ "step": 60
386
+ },
387
+ {
388
+ "epoch": 0.09,
389
+ "eval_loss": 0.6198899149894714,
390
+ "eval_runtime": 75.9771,
391
+ "eval_samples_per_second": 21.441,
392
+ "eval_steps_per_second": 10.727,
393
+ "step": 60
394
+ },
395
+ {
396
+ "epoch": 0.09,
397
+ "learning_rate": 8.133333333333334e-05,
398
+ "loss": 0.7455,
399
+ "step": 61
400
+ },
401
+ {
402
+ "epoch": 0.09,
403
+ "learning_rate": 8.266666666666667e-05,
404
+ "loss": 0.7496,
405
+ "step": 62
406
+ },
407
+ {
408
+ "epoch": 0.09,
409
+ "learning_rate": 8.4e-05,
410
+ "loss": 0.702,
411
+ "step": 63
412
+ },
413
+ {
414
+ "epoch": 0.09,
415
+ "learning_rate": 8.533333333333334e-05,
416
+ "loss": 0.8785,
417
+ "step": 64
418
+ },
419
+ {
420
+ "epoch": 0.09,
421
+ "learning_rate": 8.666666666666667e-05,
422
+ "loss": 0.7184,
423
+ "step": 65
424
+ },
425
+ {
426
+ "epoch": 0.09,
427
+ "learning_rate": 8.800000000000001e-05,
428
+ "loss": 0.7102,
429
+ "step": 66
430
+ },
431
+ {
432
+ "epoch": 0.1,
433
+ "learning_rate": 8.933333333333334e-05,
434
+ "loss": 0.795,
435
+ "step": 67
436
+ },
437
+ {
438
+ "epoch": 0.1,
439
+ "learning_rate": 9.066666666666667e-05,
440
+ "loss": 0.5903,
441
+ "step": 68
442
+ },
443
+ {
444
+ "epoch": 0.1,
445
+ "learning_rate": 9.200000000000001e-05,
446
+ "loss": 0.5926,
447
+ "step": 69
448
+ },
449
+ {
450
+ "epoch": 0.1,
451
+ "learning_rate": 9.333333333333334e-05,
452
+ "loss": 0.6543,
453
+ "step": 70
454
+ },
455
+ {
456
+ "epoch": 0.1,
457
+ "learning_rate": 9.466666666666667e-05,
458
+ "loss": 0.5384,
459
+ "step": 71
460
+ },
461
+ {
462
+ "epoch": 0.1,
463
+ "learning_rate": 9.6e-05,
464
+ "loss": 0.7371,
465
+ "step": 72
466
+ },
467
+ {
468
+ "epoch": 0.1,
469
+ "learning_rate": 9.733333333333335e-05,
470
+ "loss": 0.6365,
471
+ "step": 73
472
+ },
473
+ {
474
+ "epoch": 0.11,
475
+ "learning_rate": 9.866666666666668e-05,
476
+ "loss": 0.7859,
477
+ "step": 74
478
+ },
479
+ {
480
+ "epoch": 0.11,
481
+ "learning_rate": 0.0001,
482
+ "loss": 0.5836,
483
+ "step": 75
484
+ },
485
+ {
486
+ "epoch": 0.11,
487
+ "learning_rate": 0.00010133333333333335,
488
+ "loss": 0.5666,
489
+ "step": 76
490
+ },
491
+ {
492
+ "epoch": 0.11,
493
+ "learning_rate": 0.00010266666666666666,
494
+ "loss": 0.4644,
495
+ "step": 77
496
+ },
497
+ {
498
+ "epoch": 0.11,
499
+ "learning_rate": 0.00010400000000000001,
500
+ "loss": 0.9123,
501
+ "step": 78
502
+ },
503
+ {
504
+ "epoch": 0.11,
505
+ "learning_rate": 0.00010533333333333332,
506
+ "loss": 0.7794,
507
+ "step": 79
508
+ },
509
+ {
510
+ "epoch": 0.11,
511
+ "learning_rate": 0.00010666666666666667,
512
+ "loss": 0.7857,
513
+ "step": 80
514
+ },
515
+ {
516
+ "epoch": 0.11,
517
+ "eval_loss": 0.6065900325775146,
518
+ "eval_runtime": 75.8059,
519
+ "eval_samples_per_second": 21.489,
520
+ "eval_steps_per_second": 10.751,
521
+ "step": 80
522
+ },
523
+ {
524
+ "epoch": 0.12,
525
+ "learning_rate": 0.00010800000000000001,
526
+ "loss": 0.7882,
527
+ "step": 81
528
+ },
529
+ {
530
+ "epoch": 0.12,
531
+ "learning_rate": 0.00010933333333333333,
532
+ "loss": 0.722,
533
+ "step": 82
534
+ },
535
+ {
536
+ "epoch": 0.12,
537
+ "learning_rate": 0.00011066666666666667,
538
+ "loss": 0.6808,
539
+ "step": 83
540
+ },
541
+ {
542
+ "epoch": 0.12,
543
+ "learning_rate": 0.00011200000000000001,
544
+ "loss": 0.7176,
545
+ "step": 84
546
+ },
547
+ {
548
+ "epoch": 0.12,
549
+ "learning_rate": 0.00011333333333333334,
550
+ "loss": 0.9055,
551
+ "step": 85
552
+ },
553
+ {
554
+ "epoch": 0.12,
555
+ "learning_rate": 0.00011466666666666667,
556
+ "loss": 0.6622,
557
+ "step": 86
558
+ },
559
+ {
560
+ "epoch": 0.12,
561
+ "learning_rate": 0.000116,
562
+ "loss": 0.6591,
563
+ "step": 87
564
+ },
565
+ {
566
+ "epoch": 0.13,
567
+ "learning_rate": 0.00011733333333333334,
568
+ "loss": 0.6363,
569
+ "step": 88
570
+ },
571
+ {
572
+ "epoch": 0.13,
573
+ "learning_rate": 0.00011866666666666669,
574
+ "loss": 0.747,
575
+ "step": 89
576
+ },
577
+ {
578
+ "epoch": 0.13,
579
+ "learning_rate": 0.00012,
580
+ "loss": 0.609,
581
+ "step": 90
582
+ },
583
+ {
584
+ "epoch": 0.13,
585
+ "learning_rate": 0.00012133333333333335,
586
+ "loss": 0.8201,
587
+ "step": 91
588
+ },
589
+ {
590
+ "epoch": 0.13,
591
+ "learning_rate": 0.00012266666666666668,
592
+ "loss": 0.7042,
593
+ "step": 92
594
+ },
595
+ {
596
+ "epoch": 0.13,
597
+ "learning_rate": 0.000124,
598
+ "loss": 0.5947,
599
+ "step": 93
600
+ },
601
+ {
602
+ "epoch": 0.13,
603
+ "learning_rate": 0.00012533333333333334,
604
+ "loss": 0.8148,
605
+ "step": 94
606
+ },
607
+ {
608
+ "epoch": 0.14,
609
+ "learning_rate": 0.00012666666666666666,
610
+ "loss": 0.5936,
611
+ "step": 95
612
+ },
613
+ {
614
+ "epoch": 0.14,
615
+ "learning_rate": 0.00012800000000000002,
616
+ "loss": 0.7983,
617
+ "step": 96
618
+ },
619
+ {
620
+ "epoch": 0.14,
621
+ "learning_rate": 0.00012933333333333332,
622
+ "loss": 0.907,
623
+ "step": 97
624
+ },
625
+ {
626
+ "epoch": 0.14,
627
+ "learning_rate": 0.00013066666666666668,
628
+ "loss": 0.8807,
629
+ "step": 98
630
+ },
631
+ {
632
+ "epoch": 0.14,
633
+ "learning_rate": 0.000132,
634
+ "loss": 0.626,
635
+ "step": 99
636
+ },
637
+ {
638
+ "epoch": 0.14,
639
+ "learning_rate": 0.00013333333333333334,
640
+ "loss": 0.6208,
641
+ "step": 100
642
+ },
643
+ {
644
+ "epoch": 0.14,
645
+ "eval_loss": 0.6062822937965393,
646
+ "eval_runtime": 76.0845,
647
+ "eval_samples_per_second": 21.41,
648
+ "eval_steps_per_second": 10.712,
649
+ "step": 100
650
+ },
651
+ {
652
+ "epoch": 0.14,
653
+ "learning_rate": 0.00013466666666666667,
654
+ "loss": 0.5916,
655
+ "step": 101
656
+ },
657
+ {
658
+ "epoch": 0.15,
659
+ "learning_rate": 0.00013600000000000003,
660
+ "loss": 0.6769,
661
+ "step": 102
662
+ },
663
+ {
664
+ "epoch": 0.15,
665
+ "learning_rate": 0.00013733333333333333,
666
+ "loss": 0.6558,
667
+ "step": 103
668
+ },
669
+ {
670
+ "epoch": 0.15,
671
+ "learning_rate": 0.00013866666666666669,
672
+ "loss": 0.6152,
673
+ "step": 104
674
+ },
675
+ {
676
+ "epoch": 0.15,
677
+ "learning_rate": 0.00014,
678
+ "loss": 0.6301,
679
+ "step": 105
680
+ },
681
+ {
682
+ "epoch": 0.15,
683
+ "learning_rate": 0.00014133333333333334,
684
+ "loss": 0.6504,
685
+ "step": 106
686
+ },
687
+ {
688
+ "epoch": 0.15,
689
+ "learning_rate": 0.00014266666666666667,
690
+ "loss": 0.6567,
691
+ "step": 107
692
+ },
693
+ {
694
+ "epoch": 0.15,
695
+ "learning_rate": 0.000144,
696
+ "loss": 0.8578,
697
+ "step": 108
698
+ },
699
+ {
700
+ "epoch": 0.16,
701
+ "learning_rate": 0.00014533333333333333,
702
+ "loss": 0.7125,
703
+ "step": 109
704
+ },
705
+ {
706
+ "epoch": 0.16,
707
+ "learning_rate": 0.00014666666666666666,
708
+ "loss": 0.6192,
709
+ "step": 110
710
+ },
711
+ {
712
+ "epoch": 0.16,
713
+ "learning_rate": 0.000148,
714
+ "loss": 0.6424,
715
+ "step": 111
716
+ },
717
+ {
718
+ "epoch": 0.16,
719
+ "learning_rate": 0.00014933333333333335,
720
+ "loss": 0.5399,
721
+ "step": 112
722
+ },
723
+ {
724
+ "epoch": 0.16,
725
+ "learning_rate": 0.00015066666666666668,
726
+ "loss": 0.6844,
727
+ "step": 113
728
+ },
729
+ {
730
+ "epoch": 0.16,
731
+ "learning_rate": 0.000152,
732
+ "loss": 0.6255,
733
+ "step": 114
734
+ },
735
+ {
736
+ "epoch": 0.16,
737
+ "learning_rate": 0.00015333333333333334,
738
+ "loss": 0.7753,
739
+ "step": 115
740
+ },
741
+ {
742
+ "epoch": 0.17,
743
+ "learning_rate": 0.00015466666666666667,
744
+ "loss": 0.7796,
745
+ "step": 116
746
+ },
747
+ {
748
+ "epoch": 0.17,
749
+ "learning_rate": 0.00015600000000000002,
750
+ "loss": 0.6181,
751
+ "step": 117
752
+ },
753
+ {
754
+ "epoch": 0.17,
755
+ "learning_rate": 0.00015733333333333333,
756
+ "loss": 0.7465,
757
+ "step": 118
758
+ },
759
+ {
760
+ "epoch": 0.17,
761
+ "learning_rate": 0.00015866666666666668,
762
+ "loss": 0.7195,
763
+ "step": 119
764
+ },
765
+ {
766
+ "epoch": 0.17,
767
+ "learning_rate": 0.00016,
768
+ "loss": 0.6805,
769
+ "step": 120
770
+ },
771
+ {
772
+ "epoch": 0.17,
773
+ "eval_loss": 0.5969421863555908,
774
+ "eval_runtime": 75.9286,
775
+ "eval_samples_per_second": 21.454,
776
+ "eval_steps_per_second": 10.734,
777
+ "step": 120
778
+ },
779
+ {
780
+ "epoch": 0.17,
781
+ "learning_rate": 0.00016133333333333334,
782
+ "loss": 0.6499,
783
+ "step": 121
784
+ },
785
+ {
786
+ "epoch": 0.17,
787
+ "learning_rate": 0.00016266666666666667,
788
+ "loss": 0.5963,
789
+ "step": 122
790
+ },
791
+ {
792
+ "epoch": 0.18,
793
+ "learning_rate": 0.000164,
794
+ "loss": 0.6295,
795
+ "step": 123
796
+ },
797
+ {
798
+ "epoch": 0.18,
799
+ "learning_rate": 0.00016533333333333333,
800
+ "loss": 0.7473,
801
+ "step": 124
802
+ },
803
+ {
804
+ "epoch": 0.18,
805
+ "learning_rate": 0.0001666666666666667,
806
+ "loss": 0.65,
807
+ "step": 125
808
+ },
809
+ {
810
+ "epoch": 0.18,
811
+ "learning_rate": 0.000168,
812
+ "loss": 0.5558,
813
+ "step": 126
814
+ },
815
+ {
816
+ "epoch": 0.18,
817
+ "learning_rate": 0.00016933333333333335,
818
+ "loss": 0.5604,
819
+ "step": 127
820
+ },
821
+ {
822
+ "epoch": 0.18,
823
+ "learning_rate": 0.00017066666666666668,
824
+ "loss": 0.5784,
825
+ "step": 128
826
+ },
827
+ {
828
+ "epoch": 0.18,
829
+ "learning_rate": 0.000172,
830
+ "loss": 0.5938,
831
+ "step": 129
832
+ },
833
+ {
834
+ "epoch": 0.19,
835
+ "learning_rate": 0.00017333333333333334,
836
+ "loss": 0.8228,
837
+ "step": 130
838
+ },
839
+ {
840
+ "epoch": 0.19,
841
+ "learning_rate": 0.00017466666666666667,
842
+ "loss": 0.7301,
843
+ "step": 131
844
+ },
845
+ {
846
+ "epoch": 0.19,
847
+ "learning_rate": 0.00017600000000000002,
848
+ "loss": 0.863,
849
+ "step": 132
850
+ },
851
+ {
852
+ "epoch": 0.19,
853
+ "learning_rate": 0.00017733333333333335,
854
+ "loss": 0.5386,
855
+ "step": 133
856
+ },
857
+ {
858
+ "epoch": 0.19,
859
+ "learning_rate": 0.00017866666666666668,
860
+ "loss": 0.4405,
861
+ "step": 134
862
+ },
863
+ {
864
+ "epoch": 0.19,
865
+ "learning_rate": 0.00018,
866
+ "loss": 0.5679,
867
+ "step": 135
868
+ },
869
+ {
870
+ "epoch": 0.19,
871
+ "learning_rate": 0.00018133333333333334,
872
+ "loss": 0.718,
873
+ "step": 136
874
+ },
875
+ {
876
+ "epoch": 0.2,
877
+ "learning_rate": 0.00018266666666666667,
878
+ "loss": 0.6935,
879
+ "step": 137
880
+ },
881
+ {
882
+ "epoch": 0.2,
883
+ "learning_rate": 0.00018400000000000003,
884
+ "loss": 0.629,
885
+ "step": 138
886
+ },
887
+ {
888
+ "epoch": 0.2,
889
+ "learning_rate": 0.00018533333333333333,
890
+ "loss": 0.7571,
891
+ "step": 139
892
+ },
893
+ {
894
+ "epoch": 0.2,
895
+ "learning_rate": 0.0001866666666666667,
896
+ "loss": 0.5928,
897
+ "step": 140
898
+ },
899
+ {
900
+ "epoch": 0.2,
901
+ "eval_loss": 0.5988821983337402,
902
+ "eval_runtime": 76.0106,
903
+ "eval_samples_per_second": 21.431,
904
+ "eval_steps_per_second": 10.722,
905
+ "step": 140
906
+ },
907
+ {
908
+ "epoch": 0.2,
909
+ "learning_rate": 0.000188,
910
+ "loss": 0.4976,
911
+ "step": 141
912
+ },
913
+ {
914
+ "epoch": 0.2,
915
+ "learning_rate": 0.00018933333333333335,
916
+ "loss": 0.5056,
917
+ "step": 142
918
+ },
919
+ {
920
+ "epoch": 0.2,
921
+ "learning_rate": 0.00019066666666666668,
922
+ "loss": 0.688,
923
+ "step": 143
924
+ },
925
+ {
926
+ "epoch": 0.21,
927
+ "learning_rate": 0.000192,
928
+ "loss": 0.618,
929
+ "step": 144
930
+ },
931
+ {
932
+ "epoch": 0.21,
933
+ "learning_rate": 0.00019333333333333333,
934
+ "loss": 0.5207,
935
+ "step": 145
936
+ },
937
+ {
938
+ "epoch": 0.21,
939
+ "learning_rate": 0.0001946666666666667,
940
+ "loss": 0.45,
941
+ "step": 146
942
+ },
943
+ {
944
+ "epoch": 0.21,
945
+ "learning_rate": 0.000196,
946
+ "loss": 0.6393,
947
+ "step": 147
948
+ },
949
+ {
950
+ "epoch": 0.21,
951
+ "learning_rate": 0.00019733333333333335,
952
+ "loss": 0.6003,
953
+ "step": 148
954
+ },
955
+ {
956
+ "epoch": 0.21,
957
+ "learning_rate": 0.00019866666666666668,
958
+ "loss": 0.8397,
959
+ "step": 149
960
+ },
961
+ {
962
+ "epoch": 0.21,
963
+ "learning_rate": 0.0002,
964
+ "loss": 0.6742,
965
+ "step": 150
966
+ },
967
+ {
968
+ "epoch": 0.22,
969
+ "learning_rate": 0.00020133333333333334,
970
+ "loss": 0.5273,
971
+ "step": 151
972
+ },
973
+ {
974
+ "epoch": 0.22,
975
+ "learning_rate": 0.0002026666666666667,
976
+ "loss": 0.5878,
977
+ "step": 152
978
+ },
979
+ {
980
+ "epoch": 0.22,
981
+ "learning_rate": 0.00020400000000000003,
982
+ "loss": 0.6262,
983
+ "step": 153
984
+ },
985
+ {
986
+ "epoch": 0.22,
987
+ "learning_rate": 0.00020533333333333333,
988
+ "loss": 0.9558,
989
+ "step": 154
990
+ },
991
+ {
992
+ "epoch": 0.22,
993
+ "learning_rate": 0.00020666666666666668,
994
+ "loss": 0.563,
995
+ "step": 155
996
+ },
997
+ {
998
+ "epoch": 0.22,
999
+ "learning_rate": 0.00020800000000000001,
1000
+ "loss": 0.4969,
1001
+ "step": 156
1002
+ },
1003
+ {
1004
+ "epoch": 0.22,
1005
+ "learning_rate": 0.00020933333333333334,
1006
+ "loss": 0.6478,
1007
+ "step": 157
1008
+ },
1009
+ {
1010
+ "epoch": 0.23,
1011
+ "learning_rate": 0.00021066666666666665,
1012
+ "loss": 0.67,
1013
+ "step": 158
1014
+ },
1015
+ {
1016
+ "epoch": 0.23,
1017
+ "learning_rate": 0.00021200000000000003,
1018
+ "loss": 0.6035,
1019
+ "step": 159
1020
+ },
1021
+ {
1022
+ "epoch": 0.23,
1023
+ "learning_rate": 0.00021333333333333333,
1024
+ "loss": 0.715,
1025
+ "step": 160
1026
+ },
1027
+ {
1028
+ "epoch": 0.23,
1029
+ "eval_loss": 0.5843603014945984,
1030
+ "eval_runtime": 76.0289,
1031
+ "eval_samples_per_second": 21.426,
1032
+ "eval_steps_per_second": 10.72,
1033
+ "step": 160
1034
+ },
1035
+ {
1036
+ "epoch": 0.23,
1037
+ "learning_rate": 0.00021466666666666666,
1038
+ "loss": 0.5748,
1039
+ "step": 161
1040
+ },
1041
+ {
1042
+ "epoch": 0.23,
1043
+ "learning_rate": 0.00021600000000000002,
1044
+ "loss": 0.5162,
1045
+ "step": 162
1046
+ },
1047
+ {
1048
+ "epoch": 0.23,
1049
+ "learning_rate": 0.00021733333333333335,
1050
+ "loss": 0.4678,
1051
+ "step": 163
1052
+ },
1053
+ {
1054
+ "epoch": 0.23,
1055
+ "learning_rate": 0.00021866666666666665,
1056
+ "loss": 0.5409,
1057
+ "step": 164
1058
+ },
1059
+ {
1060
+ "epoch": 0.24,
1061
+ "learning_rate": 0.00022000000000000003,
1062
+ "loss": 0.6638,
1063
+ "step": 165
1064
+ },
1065
+ {
1066
+ "epoch": 0.24,
1067
+ "learning_rate": 0.00022133333333333334,
1068
+ "loss": 0.6777,
1069
+ "step": 166
1070
+ },
1071
+ {
1072
+ "epoch": 0.24,
1073
+ "learning_rate": 0.00022266666666666667,
1074
+ "loss": 0.5909,
1075
+ "step": 167
1076
+ },
1077
+ {
1078
+ "epoch": 0.24,
1079
+ "learning_rate": 0.00022400000000000002,
1080
+ "loss": 0.767,
1081
+ "step": 168
1082
+ },
1083
+ {
1084
+ "epoch": 0.24,
1085
+ "learning_rate": 0.00022533333333333335,
1086
+ "loss": 0.7222,
1087
+ "step": 169
1088
+ },
1089
+ {
1090
+ "epoch": 0.24,
1091
+ "learning_rate": 0.00022666666666666668,
1092
+ "loss": 0.4992,
1093
+ "step": 170
1094
+ },
1095
+ {
1096
+ "epoch": 0.24,
1097
+ "learning_rate": 0.00022799999999999999,
1098
+ "loss": 0.5737,
1099
+ "step": 171
1100
+ },
1101
+ {
1102
+ "epoch": 0.25,
1103
+ "learning_rate": 0.00022933333333333334,
1104
+ "loss": 0.4818,
1105
+ "step": 172
1106
+ },
1107
+ {
1108
+ "epoch": 0.25,
1109
+ "learning_rate": 0.00023066666666666667,
1110
+ "loss": 0.5393,
1111
+ "step": 173
1112
+ },
1113
+ {
1114
+ "epoch": 0.25,
1115
+ "learning_rate": 0.000232,
1116
+ "loss": 0.4846,
1117
+ "step": 174
1118
+ },
1119
+ {
1120
+ "epoch": 0.25,
1121
+ "learning_rate": 0.00023333333333333336,
1122
+ "loss": 0.5463,
1123
+ "step": 175
1124
+ },
1125
+ {
1126
+ "epoch": 0.25,
1127
+ "learning_rate": 0.0002346666666666667,
1128
+ "loss": 0.7732,
1129
+ "step": 176
1130
+ },
1131
+ {
1132
+ "epoch": 0.25,
1133
+ "learning_rate": 0.000236,
1134
+ "loss": 0.723,
1135
+ "step": 177
1136
+ },
1137
+ {
1138
+ "epoch": 0.25,
1139
+ "learning_rate": 0.00023733333333333337,
1140
+ "loss": 0.5236,
1141
+ "step": 178
1142
+ },
1143
+ {
1144
+ "epoch": 0.26,
1145
+ "learning_rate": 0.00023866666666666668,
1146
+ "loss": 0.5844,
1147
+ "step": 179
1148
+ },
1149
+ {
1150
+ "epoch": 0.26,
1151
+ "learning_rate": 0.00024,
1152
+ "loss": 0.5647,
1153
+ "step": 180
1154
+ },
1155
+ {
1156
+ "epoch": 0.26,
1157
+ "eval_loss": 0.5979331731796265,
1158
+ "eval_runtime": 75.9556,
1159
+ "eval_samples_per_second": 21.447,
1160
+ "eval_steps_per_second": 10.73,
1161
+ "step": 180
1162
+ },
1163
+ {
1164
+ "epoch": 0.26,
1165
+ "learning_rate": 0.00024133333333333336,
1166
+ "loss": 0.5554,
1167
+ "step": 181
1168
+ },
1169
+ {
1170
+ "epoch": 0.26,
1171
+ "learning_rate": 0.0002426666666666667,
1172
+ "loss": 0.6322,
1173
+ "step": 182
1174
+ },
1175
+ {
1176
+ "epoch": 0.26,
1177
+ "learning_rate": 0.000244,
1178
+ "loss": 0.6285,
1179
+ "step": 183
1180
+ },
1181
+ {
1182
+ "epoch": 0.26,
1183
+ "learning_rate": 0.00024533333333333335,
1184
+ "loss": 0.6452,
1185
+ "step": 184
1186
+ },
1187
+ {
1188
+ "epoch": 0.26,
1189
+ "learning_rate": 0.0002466666666666667,
1190
+ "loss": 0.663,
1191
+ "step": 185
1192
+ },
1193
+ {
1194
+ "epoch": 0.27,
1195
+ "learning_rate": 0.000248,
1196
+ "loss": 0.4447,
1197
+ "step": 186
1198
+ },
1199
+ {
1200
+ "epoch": 0.27,
1201
+ "learning_rate": 0.00024933333333333334,
1202
+ "loss": 0.5616,
1203
+ "step": 187
1204
+ },
1205
+ {
1206
+ "epoch": 0.27,
1207
+ "learning_rate": 0.00025066666666666667,
1208
+ "loss": 0.6607,
1209
+ "step": 188
1210
+ },
1211
+ {
1212
+ "epoch": 0.27,
1213
+ "learning_rate": 0.000252,
1214
+ "loss": 0.5731,
1215
+ "step": 189
1216
+ },
1217
+ {
1218
+ "epoch": 0.27,
1219
+ "learning_rate": 0.00025333333333333333,
1220
+ "loss": 0.8153,
1221
+ "step": 190
1222
+ },
1223
+ {
1224
+ "epoch": 0.27,
1225
+ "learning_rate": 0.0002546666666666667,
1226
+ "loss": 0.7403,
1227
+ "step": 191
1228
+ },
1229
+ {
1230
+ "epoch": 0.27,
1231
+ "learning_rate": 0.00025600000000000004,
1232
+ "loss": 0.5603,
1233
+ "step": 192
1234
+ },
1235
+ {
1236
+ "epoch": 0.28,
1237
+ "learning_rate": 0.0002573333333333333,
1238
+ "loss": 0.5311,
1239
+ "step": 193
1240
+ },
1241
+ {
1242
+ "epoch": 0.28,
1243
+ "learning_rate": 0.00025866666666666665,
1244
+ "loss": 0.8732,
1245
+ "step": 194
1246
+ },
1247
+ {
1248
+ "epoch": 0.28,
1249
+ "learning_rate": 0.00026000000000000003,
1250
+ "loss": 0.627,
1251
+ "step": 195
1252
+ },
1253
+ {
1254
+ "epoch": 0.28,
1255
+ "learning_rate": 0.00026133333333333336,
1256
+ "loss": 0.9056,
1257
+ "step": 196
1258
+ },
1259
+ {
1260
+ "epoch": 0.28,
1261
+ "learning_rate": 0.00026266666666666664,
1262
+ "loss": 0.4675,
1263
+ "step": 197
1264
+ },
1265
+ {
1266
+ "epoch": 0.28,
1267
+ "learning_rate": 0.000264,
1268
+ "loss": 0.5102,
1269
+ "step": 198
1270
+ },
1271
+ {
1272
+ "epoch": 0.28,
1273
+ "learning_rate": 0.00026533333333333335,
1274
+ "loss": 0.6037,
1275
+ "step": 199
1276
+ },
1277
+ {
1278
+ "epoch": 0.29,
1279
+ "learning_rate": 0.0002666666666666667,
1280
+ "loss": 0.6778,
1281
+ "step": 200
1282
+ },
1283
+ {
1284
+ "epoch": 0.29,
1285
+ "eval_loss": 0.5888569355010986,
1286
+ "eval_runtime": 75.9422,
1287
+ "eval_samples_per_second": 21.451,
1288
+ "eval_steps_per_second": 10.732,
1289
+ "step": 200
1290
+ },
1291
+ {
1292
+ "epoch": 0.29,
1293
+ "learning_rate": 0.000268,
1294
+ "loss": 0.6108,
1295
+ "step": 201
1296
+ },
1297
+ {
1298
+ "epoch": 0.29,
1299
+ "learning_rate": 0.00026933333333333334,
1300
+ "loss": 0.5723,
1301
+ "step": 202
1302
+ },
1303
+ {
1304
+ "epoch": 0.29,
1305
+ "learning_rate": 0.00027066666666666667,
1306
+ "loss": 0.6239,
1307
+ "step": 203
1308
+ },
1309
+ {
1310
+ "epoch": 0.29,
1311
+ "learning_rate": 0.00027200000000000005,
1312
+ "loss": 0.6304,
1313
+ "step": 204
1314
+ },
1315
+ {
1316
+ "epoch": 0.29,
1317
+ "learning_rate": 0.00027333333333333333,
1318
+ "loss": 0.7902,
1319
+ "step": 205
1320
+ },
1321
+ {
1322
+ "epoch": 0.29,
1323
+ "learning_rate": 0.00027466666666666666,
1324
+ "loss": 0.5582,
1325
+ "step": 206
1326
+ },
1327
+ {
1328
+ "epoch": 0.3,
1329
+ "learning_rate": 0.000276,
1330
+ "loss": 0.5657,
1331
+ "step": 207
1332
+ },
1333
+ {
1334
+ "epoch": 0.3,
1335
+ "learning_rate": 0.00027733333333333337,
1336
+ "loss": 0.6505,
1337
+ "step": 208
1338
+ },
1339
+ {
1340
+ "epoch": 0.3,
1341
+ "learning_rate": 0.0002786666666666667,
1342
+ "loss": 0.5841,
1343
+ "step": 209
1344
+ },
1345
+ {
1346
+ "epoch": 0.3,
1347
+ "learning_rate": 0.00028,
1348
+ "loss": 0.5383,
1349
+ "step": 210
1350
+ },
1351
+ {
1352
+ "epoch": 0.3,
1353
+ "learning_rate": 0.00028133333333333336,
1354
+ "loss": 0.4945,
1355
+ "step": 211
1356
+ },
1357
+ {
1358
+ "epoch": 0.3,
1359
+ "learning_rate": 0.0002826666666666667,
1360
+ "loss": 0.5968,
1361
+ "step": 212
1362
+ },
1363
+ {
1364
+ "epoch": 0.3,
1365
+ "learning_rate": 0.000284,
1366
+ "loss": 0.6705,
1367
+ "step": 213
1368
+ },
1369
+ {
1370
+ "epoch": 0.31,
1371
+ "learning_rate": 0.00028533333333333335,
1372
+ "loss": 0.571,
1373
+ "step": 214
1374
+ },
1375
+ {
1376
+ "epoch": 0.31,
1377
+ "learning_rate": 0.0002866666666666667,
1378
+ "loss": 0.5133,
1379
+ "step": 215
1380
+ },
1381
+ {
1382
+ "epoch": 0.31,
1383
+ "learning_rate": 0.000288,
1384
+ "loss": 0.6208,
1385
+ "step": 216
1386
+ },
1387
+ {
1388
+ "epoch": 0.31,
1389
+ "learning_rate": 0.0002893333333333334,
1390
+ "loss": 0.6128,
1391
+ "step": 217
1392
+ },
1393
+ {
1394
+ "epoch": 0.31,
1395
+ "learning_rate": 0.00029066666666666667,
1396
+ "loss": 0.7007,
1397
+ "step": 218
1398
+ },
1399
+ {
1400
+ "epoch": 0.31,
1401
+ "learning_rate": 0.000292,
1402
+ "loss": 0.5285,
1403
+ "step": 219
1404
+ },
1405
+ {
1406
+ "epoch": 0.31,
1407
+ "learning_rate": 0.0002933333333333333,
1408
+ "loss": 0.5907,
1409
+ "step": 220
1410
+ },
1411
+ {
1412
+ "epoch": 0.31,
1413
+ "eval_loss": 0.5771830081939697,
1414
+ "eval_runtime": 75.9871,
1415
+ "eval_samples_per_second": 21.438,
1416
+ "eval_steps_per_second": 10.725,
1417
+ "step": 220
1418
+ },
1419
+ {
1420
+ "epoch": 0.32,
1421
+ "learning_rate": 0.0002946666666666667,
1422
+ "loss": 0.5352,
1423
+ "step": 221
1424
+ },
1425
+ {
1426
+ "epoch": 0.32,
1427
+ "learning_rate": 0.000296,
1428
+ "loss": 0.5487,
1429
+ "step": 222
1430
+ },
1431
+ {
1432
+ "epoch": 0.32,
1433
+ "learning_rate": 0.0002973333333333333,
1434
+ "loss": 0.7642,
1435
+ "step": 223
1436
+ },
1437
+ {
1438
+ "epoch": 0.32,
1439
+ "learning_rate": 0.0002986666666666667,
1440
+ "loss": 0.4925,
1441
+ "step": 224
1442
+ },
1443
+ {
1444
+ "epoch": 0.32,
1445
+ "learning_rate": 0.00030000000000000003,
1446
+ "loss": 0.6328,
1447
+ "step": 225
1448
+ },
1449
+ {
1450
+ "epoch": 0.32,
1451
+ "learning_rate": 0.00030133333333333336,
1452
+ "loss": 0.8737,
1453
+ "step": 226
1454
+ },
1455
+ {
1456
+ "epoch": 0.32,
1457
+ "learning_rate": 0.0003026666666666667,
1458
+ "loss": 0.7009,
1459
+ "step": 227
1460
+ },
1461
+ {
1462
+ "epoch": 0.33,
1463
+ "learning_rate": 0.000304,
1464
+ "loss": 0.6476,
1465
+ "step": 228
1466
+ },
1467
+ {
1468
+ "epoch": 0.33,
1469
+ "learning_rate": 0.00030533333333333335,
1470
+ "loss": 0.5962,
1471
+ "step": 229
1472
+ },
1473
+ {
1474
+ "epoch": 0.33,
1475
+ "learning_rate": 0.0003066666666666667,
1476
+ "loss": 0.4297,
1477
+ "step": 230
1478
+ },
1479
+ {
1480
+ "epoch": 0.33,
1481
+ "learning_rate": 0.000308,
1482
+ "loss": 0.4098,
1483
+ "step": 231
1484
+ },
1485
+ {
1486
+ "epoch": 0.33,
1487
+ "learning_rate": 0.00030933333333333334,
1488
+ "loss": 0.466,
1489
+ "step": 232
1490
+ },
1491
+ {
1492
+ "epoch": 0.33,
1493
+ "learning_rate": 0.00031066666666666666,
1494
+ "loss": 0.557,
1495
+ "step": 233
1496
+ },
1497
+ {
1498
+ "epoch": 0.33,
1499
+ "learning_rate": 0.00031200000000000005,
1500
+ "loss": 0.5769,
1501
+ "step": 234
1502
+ },
1503
+ {
1504
+ "epoch": 0.34,
1505
+ "learning_rate": 0.0003133333333333333,
1506
+ "loss": 0.6945,
1507
+ "step": 235
1508
+ },
1509
+ {
1510
+ "epoch": 0.34,
1511
+ "learning_rate": 0.00031466666666666665,
1512
+ "loss": 0.4355,
1513
+ "step": 236
1514
+ },
1515
+ {
1516
+ "epoch": 0.34,
1517
+ "learning_rate": 0.00031600000000000004,
1518
+ "loss": 0.6799,
1519
+ "step": 237
1520
+ },
1521
+ {
1522
+ "epoch": 0.34,
1523
+ "learning_rate": 0.00031733333333333337,
1524
+ "loss": 0.5999,
1525
+ "step": 238
1526
+ },
1527
+ {
1528
+ "epoch": 0.34,
1529
+ "learning_rate": 0.00031866666666666664,
1530
+ "loss": 0.4519,
1531
+ "step": 239
1532
+ },
1533
+ {
1534
+ "epoch": 0.34,
1535
+ "learning_rate": 0.00032,
1536
+ "loss": 0.5536,
1537
+ "step": 240
1538
+ },
1539
+ {
1540
+ "epoch": 0.34,
1541
+ "eval_loss": 0.5916756987571716,
1542
+ "eval_runtime": 75.9583,
1543
+ "eval_samples_per_second": 21.446,
1544
+ "eval_steps_per_second": 10.73,
1545
+ "step": 240
1546
+ },
1547
+ {
1548
+ "epoch": 0.34,
1549
+ "learning_rate": 0.00032133333333333336,
1550
+ "loss": 0.5568,
1551
+ "step": 241
1552
+ },
1553
+ {
1554
+ "epoch": 0.35,
1555
+ "learning_rate": 0.0003226666666666667,
1556
+ "loss": 0.6273,
1557
+ "step": 242
1558
+ },
1559
+ {
1560
+ "epoch": 0.35,
1561
+ "learning_rate": 0.000324,
1562
+ "loss": 0.4417,
1563
+ "step": 243
1564
+ },
1565
+ {
1566
+ "epoch": 0.35,
1567
+ "learning_rate": 0.00032533333333333334,
1568
+ "loss": 0.5871,
1569
+ "step": 244
1570
+ },
1571
+ {
1572
+ "epoch": 0.35,
1573
+ "learning_rate": 0.0003266666666666667,
1574
+ "loss": 0.5515,
1575
+ "step": 245
1576
+ },
1577
+ {
1578
+ "epoch": 0.35,
1579
+ "learning_rate": 0.000328,
1580
+ "loss": 0.5438,
1581
+ "step": 246
1582
+ },
1583
+ {
1584
+ "epoch": 0.35,
1585
+ "learning_rate": 0.00032933333333333333,
1586
+ "loss": 0.6791,
1587
+ "step": 247
1588
+ },
1589
+ {
1590
+ "epoch": 0.35,
1591
+ "learning_rate": 0.00033066666666666666,
1592
+ "loss": 0.5649,
1593
+ "step": 248
1594
+ },
1595
+ {
1596
+ "epoch": 0.36,
1597
+ "learning_rate": 0.000332,
1598
+ "loss": 0.8795,
1599
+ "step": 249
1600
+ },
1601
+ {
1602
+ "epoch": 0.36,
1603
+ "learning_rate": 0.0003333333333333334,
1604
+ "loss": 0.5306,
1605
+ "step": 250
1606
+ },
1607
+ {
1608
+ "epoch": 0.36,
1609
+ "learning_rate": 0.0003346666666666667,
1610
+ "loss": 0.5975,
1611
+ "step": 251
1612
+ },
1613
+ {
1614
+ "epoch": 0.36,
1615
+ "learning_rate": 0.000336,
1616
+ "loss": 0.6095,
1617
+ "step": 252
1618
+ },
1619
+ {
1620
+ "epoch": 0.36,
1621
+ "learning_rate": 0.00033733333333333337,
1622
+ "loss": 0.5929,
1623
+ "step": 253
1624
+ },
1625
+ {
1626
+ "epoch": 0.36,
1627
+ "learning_rate": 0.0003386666666666667,
1628
+ "loss": 0.581,
1629
+ "step": 254
1630
+ },
1631
+ {
1632
+ "epoch": 0.36,
1633
+ "learning_rate": 0.00034,
1634
+ "loss": 0.5226,
1635
+ "step": 255
1636
+ },
1637
+ {
1638
+ "epoch": 0.37,
1639
+ "learning_rate": 0.00034133333333333335,
1640
+ "loss": 0.6894,
1641
+ "step": 256
1642
+ },
1643
+ {
1644
+ "epoch": 0.37,
1645
+ "learning_rate": 0.0003426666666666667,
1646
+ "loss": 0.5807,
1647
+ "step": 257
1648
+ },
1649
+ {
1650
+ "epoch": 0.37,
1651
+ "learning_rate": 0.000344,
1652
+ "loss": 0.5059,
1653
+ "step": 258
1654
+ },
1655
+ {
1656
+ "epoch": 0.37,
1657
+ "learning_rate": 0.00034533333333333334,
1658
+ "loss": 0.67,
1659
+ "step": 259
1660
+ },
1661
+ {
1662
+ "epoch": 0.37,
1663
+ "learning_rate": 0.00034666666666666667,
1664
+ "loss": 0.7422,
1665
+ "step": 260
1666
+ },
1667
+ {
1668
+ "epoch": 0.37,
1669
+ "eval_loss": 0.6780916452407837,
1670
+ "eval_runtime": 75.9236,
1671
+ "eval_samples_per_second": 21.456,
1672
+ "eval_steps_per_second": 10.734,
1673
+ "step": 260
1674
+ },
1675
+ {
1676
+ "epoch": 0.37,
1677
+ "learning_rate": 0.000348,
1678
+ "loss": 0.5771,
1679
+ "step": 261
1680
+ },
1681
+ {
1682
+ "epoch": 0.37,
1683
+ "learning_rate": 0.00034933333333333333,
1684
+ "loss": 0.525,
1685
+ "step": 262
1686
+ },
1687
+ {
1688
+ "epoch": 0.38,
1689
+ "learning_rate": 0.0003506666666666667,
1690
+ "loss": 0.5917,
1691
+ "step": 263
1692
+ },
1693
+ {
1694
+ "epoch": 0.38,
1695
+ "learning_rate": 0.00035200000000000005,
1696
+ "loss": 0.5993,
1697
+ "step": 264
1698
+ },
1699
+ {
1700
+ "epoch": 0.38,
1701
+ "learning_rate": 0.0003533333333333333,
1702
+ "loss": 0.5338,
1703
+ "step": 265
1704
+ },
1705
+ {
1706
+ "epoch": 0.38,
1707
+ "learning_rate": 0.0003546666666666667,
1708
+ "loss": 0.5121,
1709
+ "step": 266
1710
+ },
1711
+ {
1712
+ "epoch": 0.38,
1713
+ "learning_rate": 0.00035600000000000003,
1714
+ "loss": 0.6388,
1715
+ "step": 267
1716
+ },
1717
+ {
1718
+ "epoch": 0.38,
1719
+ "learning_rate": 0.00035733333333333336,
1720
+ "loss": 0.764,
1721
+ "step": 268
1722
+ },
1723
+ {
1724
+ "epoch": 0.38,
1725
+ "learning_rate": 0.00035866666666666664,
1726
+ "loss": 0.5963,
1727
+ "step": 269
1728
+ },
1729
+ {
1730
+ "epoch": 0.39,
1731
+ "learning_rate": 0.00036,
1732
+ "loss": 0.565,
1733
+ "step": 270
1734
+ },
1735
+ {
1736
+ "epoch": 0.39,
1737
+ "learning_rate": 0.00036133333333333335,
1738
+ "loss": 0.6174,
1739
+ "step": 271
1740
+ },
1741
+ {
1742
+ "epoch": 0.39,
1743
+ "learning_rate": 0.0003626666666666667,
1744
+ "loss": 0.629,
1745
+ "step": 272
1746
+ },
1747
+ {
1748
+ "epoch": 0.39,
1749
+ "learning_rate": 0.000364,
1750
+ "loss": 0.5634,
1751
+ "step": 273
1752
+ },
1753
+ {
1754
+ "epoch": 0.39,
1755
+ "learning_rate": 0.00036533333333333334,
1756
+ "loss": 0.535,
1757
+ "step": 274
1758
+ },
1759
+ {
1760
+ "epoch": 0.39,
1761
+ "learning_rate": 0.00036666666666666667,
1762
+ "loss": 0.654,
1763
+ "step": 275
1764
+ },
1765
+ {
1766
+ "epoch": 0.39,
1767
+ "learning_rate": 0.00036800000000000005,
1768
+ "loss": 0.6187,
1769
+ "step": 276
1770
+ },
1771
+ {
1772
+ "epoch": 0.4,
1773
+ "learning_rate": 0.00036933333333333333,
1774
+ "loss": 0.4855,
1775
+ "step": 277
1776
+ },
1777
+ {
1778
+ "epoch": 0.4,
1779
+ "learning_rate": 0.00037066666666666666,
1780
+ "loss": 0.4726,
1781
+ "step": 278
1782
+ },
1783
+ {
1784
+ "epoch": 0.4,
1785
+ "learning_rate": 0.00037200000000000004,
1786
+ "loss": 0.5277,
1787
+ "step": 279
1788
+ },
1789
+ {
1790
+ "epoch": 0.4,
1791
+ "learning_rate": 0.0003733333333333334,
1792
+ "loss": 0.6328,
1793
+ "step": 280
1794
+ },
1795
+ {
1796
+ "epoch": 0.4,
1797
+ "eval_loss": 0.5785369277000427,
1798
+ "eval_runtime": 76.0263,
1799
+ "eval_samples_per_second": 21.427,
1800
+ "eval_steps_per_second": 10.72,
1801
+ "step": 280
1802
+ },
1803
+ {
1804
+ "epoch": 0.4,
1805
+ "learning_rate": 0.0003746666666666667,
1806
+ "loss": 0.6309,
1807
+ "step": 281
1808
+ },
1809
+ {
1810
+ "epoch": 0.4,
1811
+ "learning_rate": 0.000376,
1812
+ "loss": 0.7785,
1813
+ "step": 282
1814
+ },
1815
+ {
1816
+ "epoch": 0.4,
1817
+ "learning_rate": 0.00037733333333333336,
1818
+ "loss": 0.4615,
1819
+ "step": 283
1820
+ },
1821
+ {
1822
+ "epoch": 0.41,
1823
+ "learning_rate": 0.0003786666666666667,
1824
+ "loss": 0.7048,
1825
+ "step": 284
1826
+ },
1827
+ {
1828
+ "epoch": 0.41,
1829
+ "learning_rate": 0.00038,
1830
+ "loss": 0.8294,
1831
+ "step": 285
1832
+ },
1833
+ {
1834
+ "epoch": 0.41,
1835
+ "learning_rate": 0.00038133333333333335,
1836
+ "loss": 0.6389,
1837
+ "step": 286
1838
+ },
1839
+ {
1840
+ "epoch": 0.41,
1841
+ "learning_rate": 0.0003826666666666667,
1842
+ "loss": 0.5951,
1843
+ "step": 287
1844
+ },
1845
+ {
1846
+ "epoch": 0.41,
1847
+ "learning_rate": 0.000384,
1848
+ "loss": 0.4733,
1849
+ "step": 288
1850
+ },
1851
+ {
1852
+ "epoch": 0.41,
1853
+ "learning_rate": 0.0003853333333333334,
1854
+ "loss": 0.5251,
1855
+ "step": 289
1856
+ },
1857
+ {
1858
+ "epoch": 0.41,
1859
+ "learning_rate": 0.00038666666666666667,
1860
+ "loss": 0.5581,
1861
+ "step": 290
1862
+ },
1863
+ {
1864
+ "epoch": 0.42,
1865
+ "learning_rate": 0.000388,
1866
+ "loss": 0.5923,
1867
+ "step": 291
1868
+ },
1869
+ {
1870
+ "epoch": 0.42,
1871
+ "learning_rate": 0.0003893333333333334,
1872
+ "loss": 0.5028,
1873
+ "step": 292
1874
+ },
1875
+ {
1876
+ "epoch": 0.42,
1877
+ "learning_rate": 0.0003906666666666667,
1878
+ "loss": 0.4401,
1879
+ "step": 293
1880
+ },
1881
+ {
1882
+ "epoch": 0.42,
1883
+ "learning_rate": 0.000392,
1884
+ "loss": 0.8113,
1885
+ "step": 294
1886
+ },
1887
+ {
1888
+ "epoch": 0.42,
1889
+ "learning_rate": 0.0003933333333333333,
1890
+ "loss": 0.621,
1891
+ "step": 295
1892
+ },
1893
+ {
1894
+ "epoch": 0.42,
1895
+ "learning_rate": 0.0003946666666666667,
1896
+ "loss": 0.6678,
1897
+ "step": 296
1898
+ },
1899
+ {
1900
+ "epoch": 0.42,
1901
+ "learning_rate": 0.00039600000000000003,
1902
+ "loss": 0.5044,
1903
+ "step": 297
1904
+ },
1905
+ {
1906
+ "epoch": 0.43,
1907
+ "learning_rate": 0.00039733333333333336,
1908
+ "loss": 0.5511,
1909
+ "step": 298
1910
+ },
1911
+ {
1912
+ "epoch": 0.43,
1913
+ "learning_rate": 0.0003986666666666667,
1914
+ "loss": 0.5147,
1915
+ "step": 299
1916
+ },
1917
+ {
1918
+ "epoch": 0.43,
1919
+ "learning_rate": 0.0004,
1920
+ "loss": 0.5705,
1921
+ "step": 300
1922
+ },
1923
+ {
1924
+ "epoch": 0.43,
1925
+ "eval_loss": 0.5719897747039795,
1926
+ "eval_runtime": 76.0506,
1927
+ "eval_samples_per_second": 21.42,
1928
+ "eval_steps_per_second": 10.717,
1929
+ "step": 300
1930
+ },
1931
+ {
1932
+ "epoch": 0.43,
1933
+ "learning_rate": 0.0003999938005707171,
1934
+ "loss": 0.9111,
1935
+ "step": 301
1936
+ },
1937
+ {
1938
+ "epoch": 0.43,
1939
+ "learning_rate": 0.0003999752026671977,
1940
+ "loss": 0.5359,
1941
+ "step": 302
1942
+ },
1943
+ {
1944
+ "epoch": 0.43,
1945
+ "learning_rate": 0.0003999442074424055,
1946
+ "loss": 0.7822,
1947
+ "step": 303
1948
+ },
1949
+ {
1950
+ "epoch": 0.43,
1951
+ "learning_rate": 0.00039990081681786775,
1952
+ "loss": 0.5033,
1953
+ "step": 304
1954
+ },
1955
+ {
1956
+ "epoch": 0.44,
1957
+ "learning_rate": 0.00039984503348355536,
1958
+ "loss": 0.4551,
1959
+ "step": 305
1960
+ },
1961
+ {
1962
+ "epoch": 0.44,
1963
+ "learning_rate": 0.0003997768608977168,
1964
+ "loss": 0.6405,
1965
+ "step": 306
1966
+ },
1967
+ {
1968
+ "epoch": 0.44,
1969
+ "learning_rate": 0.00039969630328666326,
1970
+ "loss": 0.4667,
1971
+ "step": 307
1972
+ },
1973
+ {
1974
+ "epoch": 0.44,
1975
+ "learning_rate": 0.0003996033656445069,
1976
+ "loss": 0.5991,
1977
+ "step": 308
1978
+ },
1979
+ {
1980
+ "epoch": 0.44,
1981
+ "learning_rate": 0.00039949805373285115,
1982
+ "loss": 0.5846,
1983
+ "step": 309
1984
+ },
1985
+ {
1986
+ "epoch": 0.44,
1987
+ "learning_rate": 0.00039938037408043353,
1988
+ "loss": 0.6403,
1989
+ "step": 310
1990
+ },
1991
+ {
1992
+ "epoch": 0.44,
1993
+ "learning_rate": 0.00039925033398272074,
1994
+ "loss": 0.5268,
1995
+ "step": 311
1996
+ },
1997
+ {
1998
+ "epoch": 0.45,
1999
+ "learning_rate": 0.00039910794150145677,
2000
+ "loss": 0.7475,
2001
+ "step": 312
2002
+ },
2003
+ {
2004
+ "epoch": 0.45,
2005
+ "learning_rate": 0.0003989532054641627,
2006
+ "loss": 0.5654,
2007
+ "step": 313
2008
+ },
2009
+ {
2010
+ "epoch": 0.45,
2011
+ "learning_rate": 0.00039878613546358993,
2012
+ "loss": 0.601,
2013
+ "step": 314
2014
+ },
2015
+ {
2016
+ "epoch": 0.45,
2017
+ "learning_rate": 0.00039860674185712486,
2018
+ "loss": 0.5911,
2019
+ "step": 315
2020
+ },
2021
+ {
2022
+ "epoch": 0.45,
2023
+ "learning_rate": 0.0003984150357661472,
2024
+ "loss": 0.5018,
2025
+ "step": 316
2026
+ },
2027
+ {
2028
+ "epoch": 0.45,
2029
+ "learning_rate": 0.0003982110290753407,
2030
+ "loss": 0.6736,
2031
+ "step": 317
2032
+ },
2033
+ {
2034
+ "epoch": 0.45,
2035
+ "learning_rate": 0.0003979947344319557,
2036
+ "loss": 0.5516,
2037
+ "step": 318
2038
+ },
2039
+ {
2040
+ "epoch": 0.46,
2041
+ "learning_rate": 0.0003977661652450257,
2042
+ "loss": 0.5712,
2043
+ "step": 319
2044
+ },
2045
+ {
2046
+ "epoch": 0.46,
2047
+ "learning_rate": 0.0003975253356845359,
2048
+ "loss": 0.6124,
2049
+ "step": 320
2050
+ },
2051
+ {
2052
+ "epoch": 0.46,
2053
+ "eval_loss": 0.5753471255302429,
2054
+ "eval_runtime": 75.9905,
2055
+ "eval_samples_per_second": 21.437,
2056
+ "eval_steps_per_second": 10.725,
2057
+ "step": 320
2058
+ },
2059
+ {
2060
+ "epoch": 0.46,
2061
+ "learning_rate": 0.0003972722606805445,
2062
+ "loss": 0.6516,
2063
+ "step": 321
2064
+ },
2065
+ {
2066
+ "epoch": 0.46,
2067
+ "learning_rate": 0.0003970069559222574,
2068
+ "loss": 0.8156,
2069
+ "step": 322
2070
+ },
2071
+ {
2072
+ "epoch": 0.46,
2073
+ "learning_rate": 0.00039672943785705553,
2074
+ "loss": 0.6279,
2075
+ "step": 323
2076
+ },
2077
+ {
2078
+ "epoch": 0.46,
2079
+ "learning_rate": 0.0003964397236894751,
2080
+ "loss": 0.7026,
2081
+ "step": 324
2082
+ },
2083
+ {
2084
+ "epoch": 0.46,
2085
+ "learning_rate": 0.00039613783138014097,
2086
+ "loss": 0.5975,
2087
+ "step": 325
2088
+ },
2089
+ {
2090
+ "epoch": 0.47,
2091
+ "learning_rate": 0.0003958237796446534,
2092
+ "loss": 0.5961,
2093
+ "step": 326
2094
+ },
2095
+ {
2096
+ "epoch": 0.47,
2097
+ "learning_rate": 0.0003954975879524277,
2098
+ "loss": 0.5417,
2099
+ "step": 327
2100
+ },
2101
+ {
2102
+ "epoch": 0.47,
2103
+ "learning_rate": 0.00039515927652548715,
2104
+ "loss": 0.7113,
2105
+ "step": 328
2106
+ },
2107
+ {
2108
+ "epoch": 0.47,
2109
+ "learning_rate": 0.00039480886633720935,
2110
+ "loss": 0.5137,
2111
+ "step": 329
2112
+ },
2113
+ {
2114
+ "epoch": 0.47,
2115
+ "learning_rate": 0.0003944463791110261,
2116
+ "loss": 0.5216,
2117
+ "step": 330
2118
+ },
2119
+ {
2120
+ "epoch": 0.47,
2121
+ "learning_rate": 0.0003940718373190768,
2122
+ "loss": 0.4807,
2123
+ "step": 331
2124
+ },
2125
+ {
2126
+ "epoch": 0.47,
2127
+ "learning_rate": 0.0003936852641808148,
2128
+ "loss": 0.5484,
2129
+ "step": 332
2130
+ },
2131
+ {
2132
+ "epoch": 0.48,
2133
+ "learning_rate": 0.00039328668366156856,
2134
+ "loss": 0.523,
2135
+ "step": 333
2136
+ },
2137
+ {
2138
+ "epoch": 0.48,
2139
+ "learning_rate": 0.0003928761204710554,
2140
+ "loss": 0.6463,
2141
+ "step": 334
2142
+ },
2143
+ {
2144
+ "epoch": 0.48,
2145
+ "learning_rate": 0.00039245360006185005,
2146
+ "loss": 0.6086,
2147
+ "step": 335
2148
+ },
2149
+ {
2150
+ "epoch": 0.48,
2151
+ "learning_rate": 0.00039201914862780653,
2152
+ "loss": 0.6831,
2153
+ "step": 336
2154
+ },
2155
+ {
2156
+ "epoch": 0.48,
2157
+ "learning_rate": 0.0003915727931024342,
2158
+ "loss": 0.446,
2159
+ "step": 337
2160
+ },
2161
+ {
2162
+ "epoch": 0.48,
2163
+ "learning_rate": 0.0003911145611572282,
2164
+ "loss": 0.73,
2165
+ "step": 338
2166
+ },
2167
+ {
2168
+ "epoch": 0.48,
2169
+ "learning_rate": 0.00039064448119995385,
2170
+ "loss": 0.5218,
2171
+ "step": 339
2172
+ },
2173
+ {
2174
+ "epoch": 0.49,
2175
+ "learning_rate": 0.00039016258237288586,
2176
+ "loss": 0.4735,
2177
+ "step": 340
2178
+ },
2179
+ {
2180
+ "epoch": 0.49,
2181
+ "eval_loss": 0.6202576160430908,
2182
+ "eval_runtime": 75.948,
2183
+ "eval_samples_per_second": 21.449,
2184
+ "eval_steps_per_second": 10.731,
2185
+ "step": 340
2186
+ },
2187
+ {
2188
+ "epoch": 0.49,
2189
+ "learning_rate": 0.0003896688945510011,
2190
+ "loss": 0.6596,
2191
+ "step": 341
2192
+ },
2193
+ {
2194
+ "epoch": 0.49,
2195
+ "learning_rate": 0.00038916344834012695,
2196
+ "loss": 0.4636,
2197
+ "step": 342
2198
+ },
2199
+ {
2200
+ "epoch": 0.49,
2201
+ "learning_rate": 0.0003886462750750439,
2202
+ "loss": 0.6812,
2203
+ "step": 343
2204
+ },
2205
+ {
2206
+ "epoch": 0.49,
2207
+ "learning_rate": 0.00038811740681754276,
2208
+ "loss": 0.435,
2209
+ "step": 344
2210
+ },
2211
+ {
2212
+ "epoch": 0.49,
2213
+ "learning_rate": 0.0003875768763544371,
2214
+ "loss": 0.6217,
2215
+ "step": 345
2216
+ },
2217
+ {
2218
+ "epoch": 0.49,
2219
+ "learning_rate": 0.0003870247171955308,
2220
+ "loss": 0.6063,
2221
+ "step": 346
2222
+ },
2223
+ {
2224
+ "epoch": 0.5,
2225
+ "learning_rate": 0.00038646096357154036,
2226
+ "loss": 0.8118,
2227
+ "step": 347
2228
+ },
2229
+ {
2230
+ "epoch": 0.5,
2231
+ "learning_rate": 0.0003858856504319732,
2232
+ "loss": 0.6175,
2233
+ "step": 348
2234
+ },
2235
+ {
2236
+ "epoch": 0.5,
2237
+ "learning_rate": 0.00038529881344296035,
2238
+ "loss": 0.6884,
2239
+ "step": 349
2240
+ },
2241
+ {
2242
+ "epoch": 0.5,
2243
+ "learning_rate": 0.0003847004889850461,
2244
+ "loss": 0.6609,
2245
+ "step": 350
2246
+ },
2247
+ {
2248
+ "epoch": 0.5,
2249
+ "learning_rate": 0.00038409071415093204,
2250
+ "loss": 0.6236,
2251
+ "step": 351
2252
+ },
2253
+ {
2254
+ "epoch": 0.5,
2255
+ "learning_rate": 0.00038346952674317783,
2256
+ "loss": 0.3354,
2257
+ "step": 352
2258
+ },
2259
+ {
2260
+ "epoch": 0.51,
2261
+ "learning_rate": 0.00038283696527185746,
2262
+ "loss": 0.6189,
2263
+ "step": 353
2264
+ },
2265
+ {
2266
+ "epoch": 0.51,
2267
+ "learning_rate": 0.0003821930689521721,
2268
+ "loss": 0.4758,
2269
+ "step": 354
2270
+ },
2271
+ {
2272
+ "epoch": 0.51,
2273
+ "learning_rate": 0.00038153787770201875,
2274
+ "loss": 0.6045,
2275
+ "step": 355
2276
+ },
2277
+ {
2278
+ "epoch": 0.51,
2279
+ "learning_rate": 0.00038087143213951554,
2280
+ "loss": 0.7085,
2281
+ "step": 356
2282
+ },
2283
+ {
2284
+ "epoch": 0.51,
2285
+ "learning_rate": 0.0003801937735804838,
2286
+ "loss": 0.8512,
2287
+ "step": 357
2288
+ },
2289
+ {
2290
+ "epoch": 0.51,
2291
+ "learning_rate": 0.0003795049440358868,
2292
+ "loss": 0.6723,
2293
+ "step": 358
2294
+ },
2295
+ {
2296
+ "epoch": 0.51,
2297
+ "learning_rate": 0.00037880498620922507,
2298
+ "loss": 0.5035,
2299
+ "step": 359
2300
+ },
2301
+ {
2302
+ "epoch": 0.52,
2303
+ "learning_rate": 0.00037809394349388895,
2304
+ "loss": 0.4602,
2305
+ "step": 360
2306
+ },
2307
+ {
2308
+ "epoch": 0.52,
2309
+ "eval_loss": 0.5771754384040833,
2310
+ "eval_runtime": 75.9251,
2311
+ "eval_samples_per_second": 21.455,
2312
+ "eval_steps_per_second": 10.734,
2313
+ "step": 360
2314
+ },
2315
+ {
2316
+ "epoch": 0.52,
2317
+ "learning_rate": 0.00037737185997046874,
2318
+ "loss": 0.5802,
2319
+ "step": 361
2320
+ },
2321
+ {
2322
+ "epoch": 0.52,
2323
+ "learning_rate": 0.00037663878040402194,
2324
+ "loss": 0.5308,
2325
+ "step": 362
2326
+ },
2327
+ {
2328
+ "epoch": 0.52,
2329
+ "learning_rate": 0.0003758947502412978,
2330
+ "loss": 0.527,
2331
+ "step": 363
2332
+ },
2333
+ {
2334
+ "epoch": 0.52,
2335
+ "learning_rate": 0.00037513981560792015,
2336
+ "loss": 0.6222,
2337
+ "step": 364
2338
+ },
2339
+ {
2340
+ "epoch": 0.52,
2341
+ "learning_rate": 0.00037437402330552766,
2342
+ "loss": 0.544,
2343
+ "step": 365
2344
+ },
2345
+ {
2346
+ "epoch": 0.52,
2347
+ "learning_rate": 0.00037359742080887265,
2348
+ "loss": 0.6717,
2349
+ "step": 366
2350
+ },
2351
+ {
2352
+ "epoch": 0.53,
2353
+ "learning_rate": 0.0003728100562628776,
2354
+ "loss": 0.5324,
2355
+ "step": 367
2356
+ },
2357
+ {
2358
+ "epoch": 0.53,
2359
+ "learning_rate": 0.00037201197847965076,
2360
+ "loss": 0.4345,
2361
+ "step": 368
2362
+ },
2363
+ {
2364
+ "epoch": 0.53,
2365
+ "learning_rate": 0.00037120323693546004,
2366
+ "loss": 0.7623,
2367
+ "step": 369
2368
+ },
2369
+ {
2370
+ "epoch": 0.53,
2371
+ "learning_rate": 0.00037038388176766546,
2372
+ "loss": 0.6668,
2373
+ "step": 370
2374
+ },
2375
+ {
2376
+ "epoch": 0.53,
2377
+ "learning_rate": 0.0003695539637716112,
2378
+ "loss": 0.6174,
2379
+ "step": 371
2380
+ },
2381
+ {
2382
+ "epoch": 0.53,
2383
+ "learning_rate": 0.0003687135343974766,
2384
+ "loss": 0.388,
2385
+ "step": 372
2386
+ },
2387
+ {
2388
+ "epoch": 0.53,
2389
+ "learning_rate": 0.0003678626457470863,
2390
+ "loss": 0.7039,
2391
+ "step": 373
2392
+ },
2393
+ {
2394
+ "epoch": 0.54,
2395
+ "learning_rate": 0.0003670013505706805,
2396
+ "loss": 0.5392,
2397
+ "step": 374
2398
+ },
2399
+ {
2400
+ "epoch": 0.54,
2401
+ "learning_rate": 0.0003661297022636446,
2402
+ "loss": 0.5504,
2403
+ "step": 375
2404
+ },
2405
+ {
2406
+ "epoch": 0.54,
2407
+ "learning_rate": 0.000365247754863199,
2408
+ "loss": 0.7533,
2409
+ "step": 376
2410
+ },
2411
+ {
2412
+ "epoch": 0.54,
2413
+ "learning_rate": 0.000364355563045049,
2414
+ "loss": 0.4881,
2415
+ "step": 377
2416
+ },
2417
+ {
2418
+ "epoch": 0.54,
2419
+ "learning_rate": 0.00036345318211999564,
2420
+ "loss": 1.0078,
2421
+ "step": 378
2422
+ },
2423
+ {
2424
+ "epoch": 0.54,
2425
+ "learning_rate": 0.00036254066803050603,
2426
+ "loss": 0.4862,
2427
+ "step": 379
2428
+ },
2429
+ {
2430
+ "epoch": 0.54,
2431
+ "learning_rate": 0.0003616180773472459,
2432
+ "loss": 0.8571,
2433
+ "step": 380
2434
+ },
2435
+ {
2436
+ "epoch": 0.54,
2437
+ "eval_loss": 0.5749652981758118,
2438
+ "eval_runtime": 75.9786,
2439
+ "eval_samples_per_second": 21.44,
2440
+ "eval_steps_per_second": 10.727,
2441
+ "step": 380
2442
+ },
2443
+ {
2444
+ "epoch": 0.55,
2445
+ "learning_rate": 0.0003606854672655723,
2446
+ "loss": 0.4636,
2447
+ "step": 381
2448
+ },
2449
+ {
2450
+ "epoch": 0.55,
2451
+ "learning_rate": 0.00035974289560198756,
2452
+ "loss": 0.6757,
2453
+ "step": 382
2454
+ },
2455
+ {
2456
+ "epoch": 0.55,
2457
+ "learning_rate": 0.00035879042079055566,
2458
+ "loss": 0.3554,
2459
+ "step": 383
2460
+ },
2461
+ {
2462
+ "epoch": 0.55,
2463
+ "learning_rate": 0.00035782810187927875,
2464
+ "loss": 0.6187,
2465
+ "step": 384
2466
+ },
2467
+ {
2468
+ "epoch": 0.55,
2469
+ "learning_rate": 0.0003568559985264373,
2470
+ "loss": 0.4611,
2471
+ "step": 385
2472
+ },
2473
+ {
2474
+ "epoch": 0.55,
2475
+ "learning_rate": 0.00035587417099689124,
2476
+ "loss": 0.4761,
2477
+ "step": 386
2478
+ },
2479
+ {
2480
+ "epoch": 0.55,
2481
+ "learning_rate": 0.0003548826801583439,
2482
+ "loss": 0.6782,
2483
+ "step": 387
2484
+ },
2485
+ {
2486
+ "epoch": 0.56,
2487
+ "learning_rate": 0.00035388158747756866,
2488
+ "loss": 0.6693,
2489
+ "step": 388
2490
+ },
2491
+ {
2492
+ "epoch": 0.56,
2493
+ "learning_rate": 0.00035287095501659844,
2494
+ "loss": 0.5747,
2495
+ "step": 389
2496
+ },
2497
+ {
2498
+ "epoch": 0.56,
2499
+ "learning_rate": 0.00035185084542887774,
2500
+ "loss": 0.6823,
2501
+ "step": 390
2502
+ },
2503
+ {
2504
+ "epoch": 0.56,
2505
+ "learning_rate": 0.00035082132195537926,
2506
+ "loss": 0.6062,
2507
+ "step": 391
2508
+ },
2509
+ {
2510
+ "epoch": 0.56,
2511
+ "learning_rate": 0.0003497824484206826,
2512
+ "loss": 0.6462,
2513
+ "step": 392
2514
+ },
2515
+ {
2516
+ "epoch": 0.56,
2517
+ "learning_rate": 0.00034873428922901797,
2518
+ "loss": 0.6891,
2519
+ "step": 393
2520
+ },
2521
+ {
2522
+ "epoch": 0.56,
2523
+ "learning_rate": 0.0003476769093602731,
2524
+ "loss": 0.5871,
2525
+ "step": 394
2526
+ },
2527
+ {
2528
+ "epoch": 0.57,
2529
+ "learning_rate": 0.0003466103743659653,
2530
+ "loss": 0.857,
2531
+ "step": 395
2532
+ },
2533
+ {
2534
+ "epoch": 0.57,
2535
+ "learning_rate": 0.00034553475036517726,
2536
+ "loss": 0.6893,
2537
+ "step": 396
2538
+ },
2539
+ {
2540
+ "epoch": 0.57,
2541
+ "learning_rate": 0.00034445010404045827,
2542
+ "loss": 0.6563,
2543
+ "step": 397
2544
+ },
2545
+ {
2546
+ "epoch": 0.57,
2547
+ "learning_rate": 0.00034335650263369027,
2548
+ "loss": 0.545,
2549
+ "step": 398
2550
+ },
2551
+ {
2552
+ "epoch": 0.57,
2553
+ "learning_rate": 0.0003422540139419191,
2554
+ "loss": 0.7041,
2555
+ "step": 399
2556
+ },
2557
+ {
2558
+ "epoch": 0.57,
2559
+ "learning_rate": 0.00034114270631315137,
2560
+ "loss": 0.5504,
2561
+ "step": 400
2562
+ },
2563
+ {
2564
+ "epoch": 0.57,
2565
+ "eval_loss": 0.6039557456970215,
2566
+ "eval_runtime": 75.7007,
2567
+ "eval_samples_per_second": 21.519,
2568
+ "eval_steps_per_second": 10.766,
2569
+ "step": 400
2570
+ },
2571
+ {
2572
+ "epoch": 0.57,
2573
+ "learning_rate": 0.00034002264864211793,
2574
+ "loss": 0.6144,
2575
+ "step": 401
2576
+ },
2577
+ {
2578
+ "epoch": 0.58,
2579
+ "learning_rate": 0.0003388939103660017,
2580
+ "loss": 0.5391,
2581
+ "step": 402
2582
+ },
2583
+ {
2584
+ "epoch": 0.58,
2585
+ "learning_rate": 0.0003377565614601343,
2586
+ "loss": 0.5719,
2587
+ "step": 403
2588
+ },
2589
+ {
2590
+ "epoch": 0.58,
2591
+ "learning_rate": 0.00033661067243365644,
2592
+ "loss": 0.4687,
2593
+ "step": 404
2594
+ },
2595
+ {
2596
+ "epoch": 0.58,
2597
+ "learning_rate": 0.00033545631432514825,
2598
+ "loss": 0.5341,
2599
+ "step": 405
2600
+ },
2601
+ {
2602
+ "epoch": 0.58,
2603
+ "learning_rate": 0.00033429355869822423,
2604
+ "loss": 0.4259,
2605
+ "step": 406
2606
+ },
2607
+ {
2608
+ "epoch": 0.58,
2609
+ "learning_rate": 0.0003331224776370972,
2610
+ "loss": 0.9081,
2611
+ "step": 407
2612
+ },
2613
+ {
2614
+ "epoch": 0.58,
2615
+ "learning_rate": 0.0003319431437421095,
2616
+ "loss": 0.4992,
2617
+ "step": 408
2618
+ },
2619
+ {
2620
+ "epoch": 0.59,
2621
+ "learning_rate": 0.00033075563012523174,
2622
+ "loss": 0.6903,
2623
+ "step": 409
2624
+ },
2625
+ {
2626
+ "epoch": 0.59,
2627
+ "learning_rate": 0.00032956001040553105,
2628
+ "loss": 0.6846,
2629
+ "step": 410
2630
+ },
2631
+ {
2632
+ "epoch": 0.59,
2633
+ "learning_rate": 0.00032835635870460635,
2634
+ "loss": 0.8808,
2635
+ "step": 411
2636
+ },
2637
+ {
2638
+ "epoch": 0.59,
2639
+ "learning_rate": 0.0003271447496419937,
2640
+ "loss": 0.7433,
2641
+ "step": 412
2642
+ },
2643
+ {
2644
+ "epoch": 0.59,
2645
+ "learning_rate": 0.0003259252583305399,
2646
+ "loss": 0.8327,
2647
+ "step": 413
2648
+ },
2649
+ {
2650
+ "epoch": 0.59,
2651
+ "learning_rate": 0.00032469796037174674,
2652
+ "loss": 0.5437,
2653
+ "step": 414
2654
+ },
2655
+ {
2656
+ "epoch": 0.59,
2657
+ "learning_rate": 0.00032346293185108306,
2658
+ "loss": 0.7766,
2659
+ "step": 415
2660
+ },
2661
+ {
2662
+ "epoch": 0.6,
2663
+ "learning_rate": 0.00032222024933326873,
2664
+ "loss": 0.5782,
2665
+ "step": 416
2666
+ },
2667
+ {
2668
+ "epoch": 0.6,
2669
+ "learning_rate": 0.00032096998985752754,
2670
+ "loss": 0.5614,
2671
+ "step": 417
2672
+ },
2673
+ {
2674
+ "epoch": 0.6,
2675
+ "learning_rate": 0.00031971223093281166,
2676
+ "loss": 0.4574,
2677
+ "step": 418
2678
+ },
2679
+ {
2680
+ "epoch": 0.6,
2681
+ "learning_rate": 0.00031844705053299605,
2682
+ "loss": 0.5042,
2683
+ "step": 419
2684
+ },
2685
+ {
2686
+ "epoch": 0.6,
2687
+ "learning_rate": 0.00031717452709204496,
2688
+ "loss": 0.6307,
2689
+ "step": 420
2690
+ },
2691
+ {
2692
+ "epoch": 0.6,
2693
+ "eval_loss": 0.579613447189331,
2694
+ "eval_runtime": 75.9225,
2695
+ "eval_samples_per_second": 21.456,
2696
+ "eval_steps_per_second": 10.735,
2697
+ "step": 420
2698
+ },
2699
+ {
2700
+ "epoch": 0.6,
2701
+ "learning_rate": 0.0003158947394991493,
2702
+ "loss": 0.5218,
2703
+ "step": 421
2704
+ },
2705
+ {
2706
+ "epoch": 0.6,
2707
+ "learning_rate": 0.0003146077670938357,
2708
+ "loss": 0.687,
2709
+ "step": 422
2710
+ },
2711
+ {
2712
+ "epoch": 0.61,
2713
+ "learning_rate": 0.0003133136896610485,
2714
+ "loss": 0.5384,
2715
+ "step": 423
2716
+ },
2717
+ {
2718
+ "epoch": 0.61,
2719
+ "learning_rate": 0.0003120125874262029,
2720
+ "loss": 0.4349,
2721
+ "step": 424
2722
+ },
2723
+ {
2724
+ "epoch": 0.61,
2725
+ "learning_rate": 0.0003107045410502118,
2726
+ "loss": 0.5447,
2727
+ "step": 425
2728
+ },
2729
+ {
2730
+ "epoch": 0.61,
2731
+ "learning_rate": 0.00030938963162448544,
2732
+ "loss": 0.4783,
2733
+ "step": 426
2734
+ },
2735
+ {
2736
+ "epoch": 0.61,
2737
+ "learning_rate": 0.0003080679406659036,
2738
+ "loss": 0.6773,
2739
+ "step": 427
2740
+ },
2741
+ {
2742
+ "epoch": 0.61,
2743
+ "learning_rate": 0.0003067395501117627,
2744
+ "loss": 0.705,
2745
+ "step": 428
2746
+ },
2747
+ {
2748
+ "epoch": 0.61,
2749
+ "learning_rate": 0.0003054045423146958,
2750
+ "loss": 0.4308,
2751
+ "step": 429
2752
+ },
2753
+ {
2754
+ "epoch": 0.62,
2755
+ "learning_rate": 0.0003040630000375671,
2756
+ "loss": 0.5592,
2757
+ "step": 430
2758
+ },
2759
+ {
2760
+ "epoch": 0.62,
2761
+ "learning_rate": 0.0003027150064483414,
2762
+ "loss": 0.7458,
2763
+ "step": 431
2764
+ },
2765
+ {
2766
+ "epoch": 0.62,
2767
+ "learning_rate": 0.00030136064511492803,
2768
+ "loss": 0.5513,
2769
+ "step": 432
2770
+ },
2771
+ {
2772
+ "epoch": 0.62,
2773
+ "learning_rate": 0.00030000000000000003,
2774
+ "loss": 0.6097,
2775
+ "step": 433
2776
+ },
2777
+ {
2778
+ "epoch": 0.62,
2779
+ "learning_rate": 0.00029863315545578916,
2780
+ "loss": 0.5079,
2781
+ "step": 434
2782
+ },
2783
+ {
2784
+ "epoch": 0.62,
2785
+ "learning_rate": 0.0002972601962188563,
2786
+ "loss": 0.6411,
2787
+ "step": 435
2788
+ },
2789
+ {
2790
+ "epoch": 0.62,
2791
+ "learning_rate": 0.00029588120740483843,
2792
+ "loss": 0.4658,
2793
+ "step": 436
2794
+ },
2795
+ {
2796
+ "epoch": 0.63,
2797
+ "learning_rate": 0.0002944962745031719,
2798
+ "loss": 0.4725,
2799
+ "step": 437
2800
+ },
2801
+ {
2802
+ "epoch": 0.63,
2803
+ "learning_rate": 0.00029310548337179267,
2804
+ "loss": 0.5727,
2805
+ "step": 438
2806
+ },
2807
+ {
2808
+ "epoch": 0.63,
2809
+ "learning_rate": 0.0002917089202318132,
2810
+ "loss": 0.4842,
2811
+ "step": 439
2812
+ },
2813
+ {
2814
+ "epoch": 0.63,
2815
+ "learning_rate": 0.0002903066716621779,
2816
+ "loss": 0.4782,
2817
+ "step": 440
2818
+ },
2819
+ {
2820
+ "epoch": 0.63,
2821
+ "eval_loss": 0.563869297504425,
2822
+ "eval_runtime": 75.946,
2823
+ "eval_samples_per_second": 21.449,
2824
+ "eval_steps_per_second": 10.731,
2825
+ "step": 440
2826
+ },
2827
+ {
2828
+ "epoch": 0.63,
2829
+ "learning_rate": 0.0002888988245942952,
2830
+ "loss": 0.5742,
2831
+ "step": 441
2832
+ },
2833
+ {
2834
+ "epoch": 0.63,
2835
+ "learning_rate": 0.0002874854663066483,
2836
+ "loss": 0.5106,
2837
+ "step": 442
2838
+ },
2839
+ {
2840
+ "epoch": 0.63,
2841
+ "learning_rate": 0.000286066684419385,
2842
+ "loss": 0.7239,
2843
+ "step": 443
2844
+ },
2845
+ {
2846
+ "epoch": 0.64,
2847
+ "learning_rate": 0.000284642566888885,
2848
+ "loss": 0.4501,
2849
+ "step": 444
2850
+ },
2851
+ {
2852
+ "epoch": 0.64,
2853
+ "learning_rate": 0.0002832132020023075,
2854
+ "loss": 0.6713,
2855
+ "step": 445
2856
+ },
2857
+ {
2858
+ "epoch": 0.64,
2859
+ "learning_rate": 0.00028177867837211775,
2860
+ "loss": 0.7934,
2861
+ "step": 446
2862
+ },
2863
+ {
2864
+ "epoch": 0.64,
2865
+ "learning_rate": 0.0002803390849305939,
2866
+ "loss": 0.6831,
2867
+ "step": 447
2868
+ },
2869
+ {
2870
+ "epoch": 0.64,
2871
+ "learning_rate": 0.0002788945109243133,
2872
+ "loss": 0.7223,
2873
+ "step": 448
2874
+ },
2875
+ {
2876
+ "epoch": 0.64,
2877
+ "learning_rate": 0.00027744504590861976,
2878
+ "loss": 0.8328,
2879
+ "step": 449
2880
+ },
2881
+ {
2882
+ "epoch": 0.64,
2883
+ "learning_rate": 0.000275990779742072,
2884
+ "loss": 0.4744,
2885
+ "step": 450
2886
+ },
2887
+ {
2888
+ "epoch": 0.65,
2889
+ "learning_rate": 0.00027453180258087264,
2890
+ "loss": 0.5313,
2891
+ "step": 451
2892
+ },
2893
+ {
2894
+ "epoch": 0.65,
2895
+ "learning_rate": 0.00027306820487327905,
2896
+ "loss": 0.4472,
2897
+ "step": 452
2898
+ },
2899
+ {
2900
+ "epoch": 0.65,
2901
+ "learning_rate": 0.00027160007735399597,
2902
+ "loss": 0.623,
2903
+ "step": 453
2904
+ },
2905
+ {
2906
+ "epoch": 0.65,
2907
+ "learning_rate": 0.00027012751103855094,
2908
+ "loss": 0.5294,
2909
+ "step": 454
2910
+ },
2911
+ {
2912
+ "epoch": 0.65,
2913
+ "learning_rate": 0.00026865059721765114,
2914
+ "loss": 0.6812,
2915
+ "step": 455
2916
+ },
2917
+ {
2918
+ "epoch": 0.65,
2919
+ "learning_rate": 0.0002671694274515246,
2920
+ "loss": 0.5081,
2921
+ "step": 456
2922
+ },
2923
+ {
2924
+ "epoch": 0.65,
2925
+ "learning_rate": 0.00026568409356424335,
2926
+ "loss": 0.6002,
2927
+ "step": 457
2928
+ },
2929
+ {
2930
+ "epoch": 0.66,
2931
+ "learning_rate": 0.00026419468763803157,
2932
+ "loss": 0.5973,
2933
+ "step": 458
2934
+ },
2935
+ {
2936
+ "epoch": 0.66,
2937
+ "learning_rate": 0.0002627013020075563,
2938
+ "loss": 0.526,
2939
+ "step": 459
2940
+ },
2941
+ {
2942
+ "epoch": 0.66,
2943
+ "learning_rate": 0.00026120402925420357,
2944
+ "loss": 0.4159,
2945
+ "step": 460
2946
+ },
2947
+ {
2948
+ "epoch": 0.66,
2949
+ "eval_loss": 0.5689454078674316,
2950
+ "eval_runtime": 76.0037,
2951
+ "eval_samples_per_second": 21.433,
2952
+ "eval_steps_per_second": 10.723,
2953
+ "step": 460
2954
+ },
2955
+ {
2956
+ "epoch": 0.66,
2957
+ "learning_rate": 0.00025970296220033894,
2958
+ "loss": 0.6363,
2959
+ "step": 461
2960
+ },
2961
+ {
2962
+ "epoch": 0.66,
2963
+ "learning_rate": 0.00025819819390355295,
2964
+ "loss": 0.4499,
2965
+ "step": 462
2966
+ },
2967
+ {
2968
+ "epoch": 0.66,
2969
+ "learning_rate": 0.000256689817650892,
2970
+ "loss": 0.4329,
2971
+ "step": 463
2972
+ },
2973
+ {
2974
+ "epoch": 0.66,
2975
+ "learning_rate": 0.00025517792695307505,
2976
+ "loss": 0.4992,
2977
+ "step": 464
2978
+ },
2979
+ {
2980
+ "epoch": 0.67,
2981
+ "learning_rate": 0.000253662615538697,
2982
+ "loss": 0.4652,
2983
+ "step": 465
2984
+ },
2985
+ {
2986
+ "epoch": 0.67,
2987
+ "learning_rate": 0.0002521439773484172,
2988
+ "loss": 0.5941,
2989
+ "step": 466
2990
+ },
2991
+ {
2992
+ "epoch": 0.67,
2993
+ "learning_rate": 0.0002506221065291364,
2994
+ "loss": 0.4763,
2995
+ "step": 467
2996
+ },
2997
+ {
2998
+ "epoch": 0.67,
2999
+ "learning_rate": 0.0002490970974281599,
3000
+ "loss": 0.5573,
3001
+ "step": 468
3002
+ },
3003
+ {
3004
+ "epoch": 0.67,
3005
+ "learning_rate": 0.0002475690445873483,
3006
+ "loss": 0.6524,
3007
+ "step": 469
3008
+ },
3009
+ {
3010
+ "epoch": 0.67,
3011
+ "learning_rate": 0.00024603804273725695,
3012
+ "loss": 0.8046,
3013
+ "step": 470
3014
+ },
3015
+ {
3016
+ "epoch": 0.67,
3017
+ "learning_rate": 0.0002445041867912629,
3018
+ "loss": 0.4634,
3019
+ "step": 471
3020
+ },
3021
+ {
3022
+ "epoch": 0.68,
3023
+ "learning_rate": 0.0002429675718396808,
3024
+ "loss": 0.9086,
3025
+ "step": 472
3026
+ },
3027
+ {
3028
+ "epoch": 0.68,
3029
+ "learning_rate": 0.00024142829314386792,
3030
+ "loss": 0.5438,
3031
+ "step": 473
3032
+ },
3033
+ {
3034
+ "epoch": 0.68,
3035
+ "learning_rate": 0.00023988644613031853,
3036
+ "loss": 0.825,
3037
+ "step": 474
3038
+ },
3039
+ {
3040
+ "epoch": 0.68,
3041
+ "learning_rate": 0.00023834212638474772,
3042
+ "loss": 0.4561,
3043
+ "step": 475
3044
+ },
3045
+ {
3046
+ "epoch": 0.68,
3047
+ "learning_rate": 0.00023679542964616618,
3048
+ "loss": 0.5544,
3049
+ "step": 476
3050
+ },
3051
+ {
3052
+ "epoch": 0.68,
3053
+ "learning_rate": 0.00023524645180094438,
3054
+ "loss": 0.5037,
3055
+ "step": 477
3056
+ },
3057
+ {
3058
+ "epoch": 0.68,
3059
+ "learning_rate": 0.00023369528887686843,
3060
+ "loss": 0.5007,
3061
+ "step": 478
3062
+ },
3063
+ {
3064
+ "epoch": 0.69,
3065
+ "learning_rate": 0.00023214203703718692,
3066
+ "loss": 0.4977,
3067
+ "step": 479
3068
+ },
3069
+ {
3070
+ "epoch": 0.69,
3071
+ "learning_rate": 0.0002305867925746492,
3072
+ "loss": 0.6393,
3073
+ "step": 480
3074
+ },
3075
+ {
3076
+ "epoch": 0.69,
3077
+ "eval_loss": 0.5660532116889954,
3078
+ "eval_runtime": 75.8556,
3079
+ "eval_samples_per_second": 21.475,
3080
+ "eval_steps_per_second": 10.744,
3081
+ "step": 480
3082
+ },
3083
+ {
3084
+ "epoch": 0.69,
3085
+ "learning_rate": 0.00022902965190553593,
3086
+ "loss": 0.8045,
3087
+ "step": 481
3088
+ },
3089
+ {
3090
+ "epoch": 0.69,
3091
+ "learning_rate": 0.00022747071156368164,
3092
+ "loss": 0.5848,
3093
+ "step": 482
3094
+ },
3095
+ {
3096
+ "epoch": 0.69,
3097
+ "learning_rate": 0.0002259100681944905,
3098
+ "loss": 0.533,
3099
+ "step": 483
3100
+ },
3101
+ {
3102
+ "epoch": 0.69,
3103
+ "learning_rate": 0.00022434781854894454,
3104
+ "loss": 0.4729,
3105
+ "step": 484
3106
+ },
3107
+ {
3108
+ "epoch": 0.69,
3109
+ "learning_rate": 0.0002227840594776056,
3110
+ "loss": 0.5992,
3111
+ "step": 485
3112
+ },
3113
+ {
3114
+ "epoch": 0.7,
3115
+ "learning_rate": 0.00022121888792461163,
3116
+ "loss": 0.4929,
3117
+ "step": 486
3118
+ },
3119
+ {
3120
+ "epoch": 0.7,
3121
+ "learning_rate": 0.00021965240092166615,
3122
+ "loss": 0.4443,
3123
+ "step": 487
3124
+ },
3125
+ {
3126
+ "epoch": 0.7,
3127
+ "learning_rate": 0.00021808469558202317,
3128
+ "loss": 0.7139,
3129
+ "step": 488
3130
+ },
3131
+ {
3132
+ "epoch": 0.7,
3133
+ "learning_rate": 0.0002165158690944665,
3134
+ "loss": 0.5218,
3135
+ "step": 489
3136
+ },
3137
+ {
3138
+ "epoch": 0.7,
3139
+ "learning_rate": 0.00021494601871728487,
3140
+ "loss": 0.5698,
3141
+ "step": 490
3142
+ },
3143
+ {
3144
+ "epoch": 0.7,
3145
+ "learning_rate": 0.00021337524177224228,
3146
+ "loss": 0.4386,
3147
+ "step": 491
3148
+ },
3149
+ {
3150
+ "epoch": 0.7,
3151
+ "learning_rate": 0.00021180363563854458,
3152
+ "loss": 0.6144,
3153
+ "step": 492
3154
+ },
3155
+ {
3156
+ "epoch": 0.71,
3157
+ "learning_rate": 0.0002102312977468027,
3158
+ "loss": 0.6238,
3159
+ "step": 493
3160
+ },
3161
+ {
3162
+ "epoch": 0.71,
3163
+ "learning_rate": 0.00020865832557299222,
3164
+ "loss": 0.5725,
3165
+ "step": 494
3166
+ },
3167
+ {
3168
+ "epoch": 0.71,
3169
+ "learning_rate": 0.00020708481663241082,
3170
+ "loss": 0.5518,
3171
+ "step": 495
3172
+ },
3173
+ {
3174
+ "epoch": 0.71,
3175
+ "learning_rate": 0.00020551086847363242,
3176
+ "loss": 0.6726,
3177
+ "step": 496
3178
+ },
3179
+ {
3180
+ "epoch": 0.71,
3181
+ "learning_rate": 0.00020393657867246018,
3182
+ "loss": 0.7597,
3183
+ "step": 497
3184
+ },
3185
+ {
3186
+ "epoch": 0.71,
3187
+ "learning_rate": 0.00020236204482587696,
3188
+ "loss": 0.7842,
3189
+ "step": 498
3190
+ },
3191
+ {
3192
+ "epoch": 0.71,
3193
+ "learning_rate": 0.0002007873645459951,
3194
+ "loss": 0.438,
3195
+ "step": 499
3196
+ },
3197
+ {
3198
+ "epoch": 0.72,
3199
+ "learning_rate": 0.000199212635454005,
3200
+ "loss": 0.8243,
3201
+ "step": 500
3202
+ },
3203
+ {
3204
+ "epoch": 0.72,
3205
+ "eval_loss": 0.5698496103286743,
3206
+ "eval_runtime": 75.9995,
3207
+ "eval_samples_per_second": 21.434,
3208
+ "eval_steps_per_second": 10.724,
3209
+ "step": 500
3210
+ },
3211
+ {
3212
+ "epoch": 0.72,
3213
+ "learning_rate": 0.0001976379551741231,
3214
+ "loss": 0.6012,
3215
+ "step": 501
3216
+ },
3217
+ {
3218
+ "epoch": 0.72,
3219
+ "learning_rate": 0.00019606342132753992,
3220
+ "loss": 0.5885,
3221
+ "step": 502
3222
+ },
3223
+ {
3224
+ "epoch": 0.72,
3225
+ "learning_rate": 0.00019448913152636762,
3226
+ "loss": 0.5198,
3227
+ "step": 503
3228
+ },
3229
+ {
3230
+ "epoch": 0.72,
3231
+ "learning_rate": 0.00019291518336758925,
3232
+ "loss": 0.555,
3233
+ "step": 504
3234
+ },
3235
+ {
3236
+ "epoch": 0.72,
3237
+ "learning_rate": 0.0001913416744270078,
3238
+ "loss": 0.5222,
3239
+ "step": 505
3240
+ },
3241
+ {
3242
+ "epoch": 0.72,
3243
+ "learning_rate": 0.0001897687022531974,
3244
+ "loss": 0.4807,
3245
+ "step": 506
3246
+ },
3247
+ {
3248
+ "epoch": 0.73,
3249
+ "learning_rate": 0.00018819636436145546,
3250
+ "loss": 0.538,
3251
+ "step": 507
3252
+ },
3253
+ {
3254
+ "epoch": 0.73,
3255
+ "learning_rate": 0.00018662475822775777,
3256
+ "loss": 0.3728,
3257
+ "step": 508
3258
+ },
3259
+ {
3260
+ "epoch": 0.73,
3261
+ "learning_rate": 0.00018505398128271515,
3262
+ "loss": 0.5586,
3263
+ "step": 509
3264
+ },
3265
+ {
3266
+ "epoch": 0.73,
3267
+ "learning_rate": 0.00018348413090553354,
3268
+ "loss": 0.821,
3269
+ "step": 510
3270
+ },
3271
+ {
3272
+ "epoch": 0.73,
3273
+ "learning_rate": 0.00018191530441797688,
3274
+ "loss": 0.7031,
3275
+ "step": 511
3276
+ },
3277
+ {
3278
+ "epoch": 0.73,
3279
+ "learning_rate": 0.00018034759907833392,
3280
+ "loss": 0.5291,
3281
+ "step": 512
3282
+ },
3283
+ {
3284
+ "epoch": 0.73,
3285
+ "learning_rate": 0.00017878111207538842,
3286
+ "loss": 0.5437,
3287
+ "step": 513
3288
+ },
3289
+ {
3290
+ "epoch": 0.74,
3291
+ "learning_rate": 0.00017721594052239444,
3292
+ "loss": 0.5258,
3293
+ "step": 514
3294
+ },
3295
+ {
3296
+ "epoch": 0.74,
3297
+ "learning_rate": 0.00017565218145105553,
3298
+ "loss": 0.9788,
3299
+ "step": 515
3300
+ },
3301
+ {
3302
+ "epoch": 0.74,
3303
+ "learning_rate": 0.00017408993180550956,
3304
+ "loss": 0.5201,
3305
+ "step": 516
3306
+ },
3307
+ {
3308
+ "epoch": 0.74,
3309
+ "learning_rate": 0.00017252928843631838,
3310
+ "loss": 0.6211,
3311
+ "step": 517
3312
+ },
3313
+ {
3314
+ "epoch": 0.74,
3315
+ "learning_rate": 0.00017097034809446414,
3316
+ "loss": 0.4348,
3317
+ "step": 518
3318
+ },
3319
+ {
3320
+ "epoch": 0.74,
3321
+ "learning_rate": 0.0001694132074253508,
3322
+ "loss": 0.465,
3323
+ "step": 519
3324
+ },
3325
+ {
3326
+ "epoch": 0.74,
3327
+ "learning_rate": 0.00016785796296281316,
3328
+ "loss": 0.4744,
3329
+ "step": 520
3330
+ },
3331
+ {
3332
+ "epoch": 0.74,
3333
+ "eval_loss": 0.5536297559738159,
3334
+ "eval_runtime": 75.93,
3335
+ "eval_samples_per_second": 21.454,
3336
+ "eval_steps_per_second": 10.734,
3337
+ "step": 520
3338
+ },
3339
+ {
3340
+ "epoch": 0.75,
3341
+ "learning_rate": 0.00016630471112313158,
3342
+ "loss": 0.5797,
3343
+ "step": 521
3344
+ },
3345
+ {
3346
+ "epoch": 0.75,
3347
+ "learning_rate": 0.0001647535481990557,
3348
+ "loss": 0.6363,
3349
+ "step": 522
3350
+ },
3351
+ {
3352
+ "epoch": 0.75,
3353
+ "learning_rate": 0.00016320457035383384,
3354
+ "loss": 0.4378,
3355
+ "step": 523
3356
+ },
3357
+ {
3358
+ "epoch": 0.75,
3359
+ "learning_rate": 0.00016165787361525238,
3360
+ "loss": 0.5703,
3361
+ "step": 524
3362
+ },
3363
+ {
3364
+ "epoch": 0.75,
3365
+ "learning_rate": 0.00016011355386968154,
3366
+ "loss": 0.3862,
3367
+ "step": 525
3368
+ },
3369
+ {
3370
+ "epoch": 0.75,
3371
+ "learning_rate": 0.00015857170685613215,
3372
+ "loss": 0.5217,
3373
+ "step": 526
3374
+ },
3375
+ {
3376
+ "epoch": 0.75,
3377
+ "learning_rate": 0.00015703242816031922,
3378
+ "loss": 0.4774,
3379
+ "step": 527
3380
+ },
3381
+ {
3382
+ "epoch": 0.76,
3383
+ "learning_rate": 0.00015549581320873715,
3384
+ "loss": 0.8355,
3385
+ "step": 528
3386
+ },
3387
+ {
3388
+ "epoch": 0.76,
3389
+ "learning_rate": 0.00015396195726274304,
3390
+ "loss": 0.4639,
3391
+ "step": 529
3392
+ },
3393
+ {
3394
+ "epoch": 0.76,
3395
+ "learning_rate": 0.0001524309554126518,
3396
+ "loss": 0.4704,
3397
+ "step": 530
3398
+ },
3399
+ {
3400
+ "epoch": 0.76,
3401
+ "learning_rate": 0.0001509029025718402,
3402
+ "loss": 0.3754,
3403
+ "step": 531
3404
+ },
3405
+ {
3406
+ "epoch": 0.76,
3407
+ "learning_rate": 0.00014937789347086364,
3408
+ "loss": 0.9537,
3409
+ "step": 532
3410
+ },
3411
+ {
3412
+ "epoch": 0.76,
3413
+ "learning_rate": 0.00014785602265158281,
3414
+ "loss": 0.4096,
3415
+ "step": 533
3416
+ },
3417
+ {
3418
+ "epoch": 0.76,
3419
+ "learning_rate": 0.00014633738446130312,
3420
+ "loss": 0.7183,
3421
+ "step": 534
3422
+ },
3423
+ {
3424
+ "epoch": 0.77,
3425
+ "learning_rate": 0.000144822073046925,
3426
+ "loss": 0.5815,
3427
+ "step": 535
3428
+ },
3429
+ {
3430
+ "epoch": 0.77,
3431
+ "learning_rate": 0.00014331018234910813,
3432
+ "loss": 0.4413,
3433
+ "step": 536
3434
+ },
3435
+ {
3436
+ "epoch": 0.77,
3437
+ "learning_rate": 0.0001418018060964471,
3438
+ "loss": 0.573,
3439
+ "step": 537
3440
+ },
3441
+ {
3442
+ "epoch": 0.77,
3443
+ "learning_rate": 0.00014029703779966114,
3444
+ "loss": 0.6439,
3445
+ "step": 538
3446
+ },
3447
+ {
3448
+ "epoch": 0.77,
3449
+ "learning_rate": 0.0001387959707457965,
3450
+ "loss": 0.4307,
3451
+ "step": 539
3452
+ },
3453
+ {
3454
+ "epoch": 0.77,
3455
+ "learning_rate": 0.00013729869799244382,
3456
+ "loss": 0.4395,
3457
+ "step": 540
3458
+ },
3459
+ {
3460
+ "epoch": 0.77,
3461
+ "eval_loss": 0.5535804629325867,
3462
+ "eval_runtime": 75.9708,
3463
+ "eval_samples_per_second": 21.442,
3464
+ "eval_steps_per_second": 10.728,
3465
+ "step": 540
3466
+ },
3467
+ {
3468
+ "epoch": 0.77,
3469
+ "learning_rate": 0.00013580531236196848,
3470
+ "loss": 0.3237,
3471
+ "step": 541
3472
+ },
3473
+ {
3474
+ "epoch": 0.78,
3475
+ "learning_rate": 0.00013431590643575673,
3476
+ "loss": 0.4944,
3477
+ "step": 542
3478
+ },
3479
+ {
3480
+ "epoch": 0.78,
3481
+ "learning_rate": 0.0001328305725484755,
3482
+ "loss": 0.4503,
3483
+ "step": 543
3484
+ },
3485
+ {
3486
+ "epoch": 0.78,
3487
+ "learning_rate": 0.00013134940278234896,
3488
+ "loss": 0.4831,
3489
+ "step": 544
3490
+ },
3491
+ {
3492
+ "epoch": 0.78,
3493
+ "learning_rate": 0.00012987248896144914,
3494
+ "loss": 0.555,
3495
+ "step": 545
3496
+ },
3497
+ {
3498
+ "epoch": 0.78,
3499
+ "learning_rate": 0.00012839992264600407,
3500
+ "loss": 0.5443,
3501
+ "step": 546
3502
+ },
3503
+ {
3504
+ "epoch": 0.78,
3505
+ "learning_rate": 0.000126931795126721,
3506
+ "loss": 0.6869,
3507
+ "step": 547
3508
+ },
3509
+ {
3510
+ "epoch": 0.78,
3511
+ "learning_rate": 0.0001254681974191274,
3512
+ "loss": 0.5072,
3513
+ "step": 548
3514
+ },
3515
+ {
3516
+ "epoch": 0.79,
3517
+ "learning_rate": 0.00012400922025792804,
3518
+ "loss": 0.5121,
3519
+ "step": 549
3520
+ },
3521
+ {
3522
+ "epoch": 0.79,
3523
+ "learning_rate": 0.00012255495409138032,
3524
+ "loss": 0.6873,
3525
+ "step": 550
3526
+ },
3527
+ {
3528
+ "epoch": 0.79,
3529
+ "learning_rate": 0.00012110548907568677,
3530
+ "loss": 0.4515,
3531
+ "step": 551
3532
+ },
3533
+ {
3534
+ "epoch": 0.79,
3535
+ "learning_rate": 0.00011966091506940616,
3536
+ "loss": 0.5488,
3537
+ "step": 552
3538
+ },
3539
+ {
3540
+ "epoch": 0.79,
3541
+ "learning_rate": 0.00011822132162788226,
3542
+ "loss": 0.5171,
3543
+ "step": 553
3544
+ },
3545
+ {
3546
+ "epoch": 0.79,
3547
+ "learning_rate": 0.00011678679799769256,
3548
+ "loss": 0.5864,
3549
+ "step": 554
3550
+ },
3551
+ {
3552
+ "epoch": 0.79,
3553
+ "learning_rate": 0.00011535743311111502,
3554
+ "loss": 0.5402,
3555
+ "step": 555
3556
+ },
3557
+ {
3558
+ "epoch": 0.8,
3559
+ "learning_rate": 0.000113933315580615,
3560
+ "loss": 0.4402,
3561
+ "step": 556
3562
+ },
3563
+ {
3564
+ "epoch": 0.8,
3565
+ "learning_rate": 0.00011251453369335172,
3566
+ "loss": 0.5805,
3567
+ "step": 557
3568
+ },
3569
+ {
3570
+ "epoch": 0.8,
3571
+ "learning_rate": 0.00011110117540570493,
3572
+ "loss": 0.8398,
3573
+ "step": 558
3574
+ },
3575
+ {
3576
+ "epoch": 0.8,
3577
+ "learning_rate": 0.00010969332833782218,
3578
+ "loss": 0.653,
3579
+ "step": 559
3580
+ },
3581
+ {
3582
+ "epoch": 0.8,
3583
+ "learning_rate": 0.00010829107976818687,
3584
+ "loss": 0.543,
3585
+ "step": 560
3586
+ },
3587
+ {
3588
+ "epoch": 0.8,
3589
+ "eval_loss": 0.5492686033248901,
3590
+ "eval_runtime": 75.8627,
3591
+ "eval_samples_per_second": 21.473,
3592
+ "eval_steps_per_second": 10.743,
3593
+ "step": 560
3594
+ },
3595
+ {
3596
+ "epoch": 0.8,
3597
+ "learning_rate": 0.00010689451662820737,
3598
+ "loss": 0.582,
3599
+ "step": 561
3600
+ },
3601
+ {
3602
+ "epoch": 0.8,
3603
+ "learning_rate": 0.00010550372549682808,
3604
+ "loss": 0.4656,
3605
+ "step": 562
3606
+ },
3607
+ {
3608
+ "epoch": 0.81,
3609
+ "learning_rate": 0.00010411879259516158,
3610
+ "loss": 0.6783,
3611
+ "step": 563
3612
+ },
3613
+ {
3614
+ "epoch": 0.81,
3615
+ "learning_rate": 0.00010273980378114374,
3616
+ "loss": 0.4788,
3617
+ "step": 564
3618
+ },
3619
+ {
3620
+ "epoch": 0.81,
3621
+ "learning_rate": 0.00010136684454421091,
3622
+ "loss": 0.4883,
3623
+ "step": 565
3624
+ },
3625
+ {
3626
+ "epoch": 0.81,
3627
+ "learning_rate": 0.00010000000000000005,
3628
+ "loss": 0.685,
3629
+ "step": 566
3630
+ },
3631
+ {
3632
+ "epoch": 0.81,
3633
+ "learning_rate": 9.863935488507201e-05,
3634
+ "loss": 0.5525,
3635
+ "step": 567
3636
+ },
3637
+ {
3638
+ "epoch": 0.81,
3639
+ "learning_rate": 9.72849935516587e-05,
3640
+ "loss": 0.3223,
3641
+ "step": 568
3642
+ },
3643
+ {
3644
+ "epoch": 0.81,
3645
+ "learning_rate": 9.593699996243295e-05,
3646
+ "loss": 0.5418,
3647
+ "step": 569
3648
+ },
3649
+ {
3650
+ "epoch": 0.82,
3651
+ "learning_rate": 9.459545768530425e-05,
3652
+ "loss": 0.5551,
3653
+ "step": 570
3654
+ },
3655
+ {
3656
+ "epoch": 0.82,
3657
+ "learning_rate": 9.326044988823733e-05,
3658
+ "loss": 0.5034,
3659
+ "step": 571
3660
+ },
3661
+ {
3662
+ "epoch": 0.82,
3663
+ "learning_rate": 9.193205933409646e-05,
3664
+ "loss": 0.3789,
3665
+ "step": 572
3666
+ },
3667
+ {
3668
+ "epoch": 0.82,
3669
+ "learning_rate": 9.061036837551466e-05,
3670
+ "loss": 0.5687,
3671
+ "step": 573
3672
+ },
3673
+ {
3674
+ "epoch": 0.82,
3675
+ "learning_rate": 8.929545894978825e-05,
3676
+ "loss": 0.3573,
3677
+ "step": 574
3678
+ },
3679
+ {
3680
+ "epoch": 0.82,
3681
+ "learning_rate": 8.798741257379714e-05,
3682
+ "loss": 0.4189,
3683
+ "step": 575
3684
+ },
3685
+ {
3686
+ "epoch": 0.82,
3687
+ "learning_rate": 8.66863103389516e-05,
3688
+ "loss": 0.4463,
3689
+ "step": 576
3690
+ },
3691
+ {
3692
+ "epoch": 0.83,
3693
+ "learning_rate": 8.53922329061643e-05,
3694
+ "loss": 0.7484,
3695
+ "step": 577
3696
+ },
3697
+ {
3698
+ "epoch": 0.83,
3699
+ "learning_rate": 8.410526050085077e-05,
3700
+ "loss": 0.5024,
3701
+ "step": 578
3702
+ },
3703
+ {
3704
+ "epoch": 0.83,
3705
+ "learning_rate": 8.282547290795509e-05,
3706
+ "loss": 0.5854,
3707
+ "step": 579
3708
+ },
3709
+ {
3710
+ "epoch": 0.83,
3711
+ "learning_rate": 8.155294946700402e-05,
3712
+ "loss": 0.4451,
3713
+ "step": 580
3714
+ },
3715
+ {
3716
+ "epoch": 0.83,
3717
+ "eval_loss": 0.5421439409255981,
3718
+ "eval_runtime": 75.9322,
3719
+ "eval_samples_per_second": 21.453,
3720
+ "eval_steps_per_second": 10.733,
3721
+ "step": 580
3722
+ },
3723
+ {
3724
+ "epoch": 0.83,
3725
+ "learning_rate": 8.02877690671884e-05,
3726
+ "loss": 0.2904,
3727
+ "step": 581
3728
+ },
3729
+ {
3730
+ "epoch": 0.83,
3731
+ "learning_rate": 7.903001014247255e-05,
3732
+ "loss": 0.5851,
3733
+ "step": 582
3734
+ },
3735
+ {
3736
+ "epoch": 0.83,
3737
+ "learning_rate": 7.777975066673133e-05,
3738
+ "loss": 0.7889,
3739
+ "step": 583
3740
+ },
3741
+ {
3742
+ "epoch": 0.84,
3743
+ "learning_rate": 7.653706814891697e-05,
3744
+ "loss": 0.4611,
3745
+ "step": 584
3746
+ },
3747
+ {
3748
+ "epoch": 0.84,
3749
+ "learning_rate": 7.530203962825331e-05,
3750
+ "loss": 0.5179,
3751
+ "step": 585
3752
+ },
3753
+ {
3754
+ "epoch": 0.84,
3755
+ "learning_rate": 7.407474166946011e-05,
3756
+ "loss": 0.5432,
3757
+ "step": 586
3758
+ },
3759
+ {
3760
+ "epoch": 0.84,
3761
+ "learning_rate": 7.285525035800644e-05,
3762
+ "loss": 0.4547,
3763
+ "step": 587
3764
+ },
3765
+ {
3766
+ "epoch": 0.84,
3767
+ "learning_rate": 7.164364129539373e-05,
3768
+ "loss": 0.8264,
3769
+ "step": 588
3770
+ },
3771
+ {
3772
+ "epoch": 0.84,
3773
+ "learning_rate": 7.043998959446896e-05,
3774
+ "loss": 0.768,
3775
+ "step": 589
3776
+ },
3777
+ {
3778
+ "epoch": 0.84,
3779
+ "learning_rate": 6.924436987476832e-05,
3780
+ "loss": 0.5146,
3781
+ "step": 590
3782
+ },
3783
+ {
3784
+ "epoch": 0.85,
3785
+ "learning_rate": 6.805685625789057e-05,
3786
+ "loss": 0.7737,
3787
+ "step": 591
3788
+ },
3789
+ {
3790
+ "epoch": 0.85,
3791
+ "learning_rate": 6.687752236290283e-05,
3792
+ "loss": 0.7187,
3793
+ "step": 592
3794
+ },
3795
+ {
3796
+ "epoch": 0.85,
3797
+ "learning_rate": 6.570644130177584e-05,
3798
+ "loss": 0.3896,
3799
+ "step": 593
3800
+ },
3801
+ {
3802
+ "epoch": 0.85,
3803
+ "learning_rate": 6.454368567485183e-05,
3804
+ "loss": 0.4378,
3805
+ "step": 594
3806
+ },
3807
+ {
3808
+ "epoch": 0.85,
3809
+ "learning_rate": 6.338932756634354e-05,
3810
+ "loss": 0.4103,
3811
+ "step": 595
3812
+ },
3813
+ {
3814
+ "epoch": 0.85,
3815
+ "learning_rate": 6.224343853986583e-05,
3816
+ "loss": 0.6346,
3817
+ "step": 596
3818
+ },
3819
+ {
3820
+ "epoch": 0.85,
3821
+ "learning_rate": 6.110608963399827e-05,
3822
+ "loss": 0.5081,
3823
+ "step": 597
3824
+ },
3825
+ {
3826
+ "epoch": 0.86,
3827
+ "learning_rate": 5.997735135788216e-05,
3828
+ "loss": 0.414,
3829
+ "step": 598
3830
+ },
3831
+ {
3832
+ "epoch": 0.86,
3833
+ "learning_rate": 5.8857293686848646e-05,
3834
+ "loss": 0.5079,
3835
+ "step": 599
3836
+ },
3837
+ {
3838
+ "epoch": 0.86,
3839
+ "learning_rate": 5.7745986058081016e-05,
3840
+ "loss": 0.5384,
3841
+ "step": 600
3842
+ },
3843
+ {
3844
+ "epoch": 0.86,
3845
+ "eval_loss": 0.5467327237129211,
3846
+ "eval_runtime": 75.8761,
3847
+ "eval_samples_per_second": 21.469,
3848
+ "eval_steps_per_second": 10.741,
3849
+ "step": 600
3850
+ },
3851
+ {
3852
+ "epoch": 0.86,
3853
+ "learning_rate": 5.6643497366309785e-05,
3854
+ "loss": 0.4515,
3855
+ "step": 601
3856
+ },
3857
+ {
3858
+ "epoch": 0.86,
3859
+ "learning_rate": 5.554989595954177e-05,
3860
+ "loss": 0.6003,
3861
+ "step": 602
3862
+ },
3863
+ {
3864
+ "epoch": 0.86,
3865
+ "learning_rate": 5.4465249634822756e-05,
3866
+ "loss": 0.7991,
3867
+ "step": 603
3868
+ },
3869
+ {
3870
+ "epoch": 0.86,
3871
+ "learning_rate": 5.338962563403478e-05,
3872
+ "loss": 0.5519,
3873
+ "step": 604
3874
+ },
3875
+ {
3876
+ "epoch": 0.87,
3877
+ "learning_rate": 5.232309063972691e-05,
3878
+ "loss": 0.5531,
3879
+ "step": 605
3880
+ },
3881
+ {
3882
+ "epoch": 0.87,
3883
+ "learning_rate": 5.126571077098208e-05,
3884
+ "loss": 0.6466,
3885
+ "step": 606
3886
+ },
3887
+ {
3888
+ "epoch": 0.87,
3889
+ "learning_rate": 5.02175515793174e-05,
3890
+ "loss": 0.4096,
3891
+ "step": 607
3892
+ },
3893
+ {
3894
+ "epoch": 0.87,
3895
+ "learning_rate": 4.917867804462077e-05,
3896
+ "loss": 0.4503,
3897
+ "step": 608
3898
+ },
3899
+ {
3900
+ "epoch": 0.87,
3901
+ "learning_rate": 4.81491545711223e-05,
3902
+ "loss": 0.4375,
3903
+ "step": 609
3904
+ },
3905
+ {
3906
+ "epoch": 0.87,
3907
+ "learning_rate": 4.712904498340167e-05,
3908
+ "loss": 0.4413,
3909
+ "step": 610
3910
+ },
3911
+ {
3912
+ "epoch": 0.87,
3913
+ "learning_rate": 4.611841252243134e-05,
3914
+ "loss": 0.5404,
3915
+ "step": 611
3916
+ },
3917
+ {
3918
+ "epoch": 0.88,
3919
+ "learning_rate": 4.511731984165619e-05,
3920
+ "loss": 0.532,
3921
+ "step": 612
3922
+ },
3923
+ {
3924
+ "epoch": 0.88,
3925
+ "learning_rate": 4.4125829003108824e-05,
3926
+ "loss": 0.5145,
3927
+ "step": 613
3928
+ },
3929
+ {
3930
+ "epoch": 0.88,
3931
+ "learning_rate": 4.3144001473562746e-05,
3932
+ "loss": 0.5592,
3933
+ "step": 614
3934
+ },
3935
+ {
3936
+ "epoch": 0.88,
3937
+ "learning_rate": 4.217189812072131e-05,
3938
+ "loss": 0.5453,
3939
+ "step": 615
3940
+ },
3941
+ {
3942
+ "epoch": 0.88,
3943
+ "learning_rate": 4.1209579209444416e-05,
3944
+ "loss": 0.5213,
3945
+ "step": 616
3946
+ },
3947
+ {
3948
+ "epoch": 0.88,
3949
+ "learning_rate": 4.02571043980124e-05,
3950
+ "loss": 0.5131,
3951
+ "step": 617
3952
+ },
3953
+ {
3954
+ "epoch": 0.88,
3955
+ "learning_rate": 3.93145327344278e-05,
3956
+ "loss": 0.5372,
3957
+ "step": 618
3958
+ },
3959
+ {
3960
+ "epoch": 0.89,
3961
+ "learning_rate": 3.838192265275411e-05,
3962
+ "loss": 0.4571,
3963
+ "step": 619
3964
+ },
3965
+ {
3966
+ "epoch": 0.89,
3967
+ "learning_rate": 3.7459331969494e-05,
3968
+ "loss": 0.4438,
3969
+ "step": 620
3970
+ },
3971
+ {
3972
+ "epoch": 0.89,
3973
+ "eval_loss": 0.5379120111465454,
3974
+ "eval_runtime": 75.9975,
3975
+ "eval_samples_per_second": 21.435,
3976
+ "eval_steps_per_second": 10.724,
3977
+ "step": 620
3978
+ },
3979
+ {
3980
+ "epoch": 0.89,
3981
+ "learning_rate": 3.65468178800044e-05,
3982
+ "loss": 0.5414,
3983
+ "step": 621
3984
+ },
3985
+ {
3986
+ "epoch": 0.89,
3987
+ "learning_rate": 3.564443695495099e-05,
3988
+ "loss": 0.5231,
3989
+ "step": 622
3990
+ },
3991
+ {
3992
+ "epoch": 0.89,
3993
+ "learning_rate": 3.4752245136801065e-05,
3994
+ "loss": 0.5973,
3995
+ "step": 623
3996
+ },
3997
+ {
3998
+ "epoch": 0.89,
3999
+ "learning_rate": 3.387029773635544e-05,
4000
+ "loss": 0.2479,
4001
+ "step": 624
4002
+ },
4003
+ {
4004
+ "epoch": 0.89,
4005
+ "learning_rate": 3.299864942931952e-05,
4006
+ "loss": 0.5817,
4007
+ "step": 625
4008
+ },
4009
+ {
4010
+ "epoch": 0.9,
4011
+ "learning_rate": 3.21373542529138e-05,
4012
+ "loss": 0.5529,
4013
+ "step": 626
4014
+ },
4015
+ {
4016
+ "epoch": 0.9,
4017
+ "learning_rate": 3.128646560252346e-05,
4018
+ "loss": 0.6231,
4019
+ "step": 627
4020
+ },
4021
+ {
4022
+ "epoch": 0.9,
4023
+ "learning_rate": 3.0446036228388842e-05,
4024
+ "loss": 0.7211,
4025
+ "step": 628
4026
+ },
4027
+ {
4028
+ "epoch": 0.9,
4029
+ "learning_rate": 2.961611823233461e-05,
4030
+ "loss": 0.4726,
4031
+ "step": 629
4032
+ },
4033
+ {
4034
+ "epoch": 0.9,
4035
+ "learning_rate": 2.879676306454e-05,
4036
+ "loss": 0.5618,
4037
+ "step": 630
4038
+ },
4039
+ {
4040
+ "epoch": 0.9,
4041
+ "learning_rate": 2.7988021520349207e-05,
4042
+ "loss": 0.4254,
4043
+ "step": 631
4044
+ },
4045
+ {
4046
+ "epoch": 0.9,
4047
+ "learning_rate": 2.718994373712247e-05,
4048
+ "loss": 0.5931,
4049
+ "step": 632
4050
+ },
4051
+ {
4052
+ "epoch": 0.91,
4053
+ "learning_rate": 2.6402579191127387e-05,
4054
+ "loss": 0.9991,
4055
+ "step": 633
4056
+ },
4057
+ {
4058
+ "epoch": 0.91,
4059
+ "learning_rate": 2.562597669447233e-05,
4060
+ "loss": 0.5177,
4061
+ "step": 634
4062
+ },
4063
+ {
4064
+ "epoch": 0.91,
4065
+ "learning_rate": 2.4860184392079865e-05,
4066
+ "loss": 0.4434,
4067
+ "step": 635
4068
+ },
4069
+ {
4070
+ "epoch": 0.91,
4071
+ "learning_rate": 2.410524975870221e-05,
4072
+ "loss": 0.4668,
4073
+ "step": 636
4074
+ },
4075
+ {
4076
+ "epoch": 0.91,
4077
+ "learning_rate": 2.336121959597808e-05,
4078
+ "loss": 0.4063,
4079
+ "step": 637
4080
+ },
4081
+ {
4082
+ "epoch": 0.91,
4083
+ "learning_rate": 2.2628140029531286e-05,
4084
+ "loss": 0.6887,
4085
+ "step": 638
4086
+ },
4087
+ {
4088
+ "epoch": 0.91,
4089
+ "learning_rate": 2.1906056506111083e-05,
4090
+ "loss": 0.3671,
4091
+ "step": 639
4092
+ },
4093
+ {
4094
+ "epoch": 0.92,
4095
+ "learning_rate": 2.1195013790775e-05,
4096
+ "loss": 0.4168,
4097
+ "step": 640
4098
+ },
4099
+ {
4100
+ "epoch": 0.92,
4101
+ "eval_loss": 0.5397519469261169,
4102
+ "eval_runtime": 75.9178,
4103
+ "eval_samples_per_second": 21.457,
4104
+ "eval_steps_per_second": 10.735,
4105
+ "step": 640
4106
+ },
4107
+ {
4108
+ "epoch": 0.92,
4109
+ "learning_rate": 2.0495055964113165e-05,
4110
+ "loss": 0.5466,
4111
+ "step": 641
4112
+ },
4113
+ {
4114
+ "epoch": 0.92,
4115
+ "learning_rate": 1.9806226419516192e-05,
4116
+ "loss": 0.5021,
4117
+ "step": 642
4118
+ },
4119
+ {
4120
+ "epoch": 0.92,
4121
+ "learning_rate": 1.9128567860484513e-05,
4122
+ "loss": 0.668,
4123
+ "step": 643
4124
+ },
4125
+ {
4126
+ "epoch": 0.92,
4127
+ "learning_rate": 1.84621222979813e-05,
4128
+ "loss": 0.6027,
4129
+ "step": 644
4130
+ },
4131
+ {
4132
+ "epoch": 0.92,
4133
+ "learning_rate": 1.7806931047827914e-05,
4134
+ "loss": 0.8838,
4135
+ "step": 645
4136
+ },
4137
+ {
4138
+ "epoch": 0.92,
4139
+ "learning_rate": 1.7163034728142537e-05,
4140
+ "loss": 0.5397,
4141
+ "step": 646
4142
+ },
4143
+ {
4144
+ "epoch": 0.93,
4145
+ "learning_rate": 1.6530473256822175e-05,
4146
+ "loss": 0.388,
4147
+ "step": 647
4148
+ },
4149
+ {
4150
+ "epoch": 0.93,
4151
+ "learning_rate": 1.590928584906799e-05,
4152
+ "loss": 0.39,
4153
+ "step": 648
4154
+ },
4155
+ {
4156
+ "epoch": 0.93,
4157
+ "learning_rate": 1.5299511014953904e-05,
4158
+ "loss": 0.4025,
4159
+ "step": 649
4160
+ },
4161
+ {
4162
+ "epoch": 0.93,
4163
+ "learning_rate": 1.4701186557039648e-05,
4164
+ "loss": 0.5292,
4165
+ "step": 650
4166
+ },
4167
+ {
4168
+ "epoch": 0.93,
4169
+ "learning_rate": 1.4114349568026863e-05,
4170
+ "loss": 0.4569,
4171
+ "step": 651
4172
+ },
4173
+ {
4174
+ "epoch": 0.93,
4175
+ "learning_rate": 1.353903642845964e-05,
4176
+ "loss": 0.3907,
4177
+ "step": 652
4178
+ },
4179
+ {
4180
+ "epoch": 0.93,
4181
+ "learning_rate": 1.2975282804469225e-05,
4182
+ "loss": 0.3466,
4183
+ "step": 653
4184
+ },
4185
+ {
4186
+ "epoch": 0.94,
4187
+ "learning_rate": 1.2423123645562951e-05,
4188
+ "loss": 0.5364,
4189
+ "step": 654
4190
+ },
4191
+ {
4192
+ "epoch": 0.94,
4193
+ "learning_rate": 1.188259318245728e-05,
4194
+ "loss": 0.5549,
4195
+ "step": 655
4196
+ },
4197
+ {
4198
+ "epoch": 0.94,
4199
+ "learning_rate": 1.1353724924956121e-05,
4200
+ "loss": 0.5845,
4201
+ "step": 656
4202
+ },
4203
+ {
4204
+ "epoch": 0.94,
4205
+ "learning_rate": 1.0836551659873074e-05,
4206
+ "loss": 0.5839,
4207
+ "step": 657
4208
+ },
4209
+ {
4210
+ "epoch": 0.94,
4211
+ "learning_rate": 1.0331105448998957e-05,
4212
+ "loss": 0.4616,
4213
+ "step": 658
4214
+ },
4215
+ {
4216
+ "epoch": 0.94,
4217
+ "learning_rate": 9.837417627114187e-06,
4218
+ "loss": 0.5385,
4219
+ "step": 659
4220
+ },
4221
+ {
4222
+ "epoch": 0.94,
4223
+ "learning_rate": 9.355518800046171e-06,
4224
+ "loss": 0.469,
4225
+ "step": 660
4226
+ },
4227
+ {
4228
+ "epoch": 0.94,
4229
+ "eval_loss": 0.5402470827102661,
4230
+ "eval_runtime": 75.9472,
4231
+ "eval_samples_per_second": 21.449,
4232
+ "eval_steps_per_second": 10.731,
4233
+ "step": 660
4234
+ },
4235
+ {
4236
+ "epoch": 0.95,
4237
+ "learning_rate": 8.885438842771843e-06,
4238
+ "loss": 0.5539,
4239
+ "step": 661
4240
+ },
4241
+ {
4242
+ "epoch": 0.95,
4243
+ "learning_rate": 8.427206897565865e-06,
4244
+ "loss": 0.313,
4245
+ "step": 662
4246
+ },
4247
+ {
4248
+ "epoch": 0.95,
4249
+ "learning_rate": 7.980851372193488e-06,
4250
+ "loss": 0.4776,
4251
+ "step": 663
4252
+ },
4253
+ {
4254
+ "epoch": 0.95,
4255
+ "learning_rate": 7.546399938149917e-06,
4256
+ "loss": 0.5348,
4257
+ "step": 664
4258
+ },
4259
+ {
4260
+ "epoch": 0.95,
4261
+ "learning_rate": 7.123879528944599e-06,
4262
+ "loss": 0.4273,
4263
+ "step": 665
4264
+ },
4265
+ {
4266
+ "epoch": 0.95,
4267
+ "learning_rate": 6.713316338431486e-06,
4268
+ "loss": 0.5493,
4269
+ "step": 666
4270
+ },
4271
+ {
4272
+ "epoch": 0.95,
4273
+ "learning_rate": 6.314735819185202e-06,
4274
+ "loss": 0.5291,
4275
+ "step": 667
4276
+ },
4277
+ {
4278
+ "epoch": 0.96,
4279
+ "learning_rate": 5.928162680923266e-06,
4280
+ "loss": 0.5378,
4281
+ "step": 668
4282
+ },
4283
+ {
4284
+ "epoch": 0.96,
4285
+ "learning_rate": 5.553620888973887e-06,
4286
+ "loss": 0.7479,
4287
+ "step": 669
4288
+ },
4289
+ {
4290
+ "epoch": 0.96,
4291
+ "learning_rate": 5.191133662790692e-06,
4292
+ "loss": 0.6171,
4293
+ "step": 670
4294
+ },
4295
+ {
4296
+ "epoch": 0.96,
4297
+ "learning_rate": 4.840723474512876e-06,
4298
+ "loss": 0.5556,
4299
+ "step": 671
4300
+ },
4301
+ {
4302
+ "epoch": 0.96,
4303
+ "learning_rate": 4.502412047572291e-06,
4304
+ "loss": 0.4105,
4305
+ "step": 672
4306
+ },
4307
+ {
4308
+ "epoch": 0.96,
4309
+ "learning_rate": 4.176220355346594e-06,
4310
+ "loss": 0.5382,
4311
+ "step": 673
4312
+ },
4313
+ {
4314
+ "epoch": 0.96,
4315
+ "learning_rate": 3.862168619859041e-06,
4316
+ "loss": 0.4685,
4317
+ "step": 674
4318
+ },
4319
+ {
4320
+ "epoch": 0.97,
4321
+ "learning_rate": 3.5602763105249125e-06,
4322
+ "loss": 0.4685,
4323
+ "step": 675
4324
+ },
4325
+ {
4326
+ "epoch": 0.97,
4327
+ "learning_rate": 3.270562142944478e-06,
4328
+ "loss": 0.5683,
4329
+ "step": 676
4330
+ },
4331
+ {
4332
+ "epoch": 0.97,
4333
+ "learning_rate": 2.9930440777425905e-06,
4334
+ "loss": 0.376,
4335
+ "step": 677
4336
+ },
4337
+ {
4338
+ "epoch": 0.97,
4339
+ "learning_rate": 2.7277393194555358e-06,
4340
+ "loss": 0.4913,
4341
+ "step": 678
4342
+ },
4343
+ {
4344
+ "epoch": 0.97,
4345
+ "learning_rate": 2.474664315464126e-06,
4346
+ "loss": 0.4781,
4347
+ "step": 679
4348
+ },
4349
+ {
4350
+ "epoch": 0.97,
4351
+ "learning_rate": 2.2338347549742956e-06,
4352
+ "loss": 0.6766,
4353
+ "step": 680
4354
+ },
4355
+ {
4356
+ "epoch": 0.97,
4357
+ "eval_loss": 0.5407062768936157,
4358
+ "eval_runtime": 76.0027,
4359
+ "eval_samples_per_second": 21.433,
4360
+ "eval_steps_per_second": 10.723,
4361
+ "step": 680
4362
+ },
4363
+ {
4364
+ "epoch": 0.97,
4365
+ "learning_rate": 2.0052655680443457e-06,
4366
+ "loss": 0.6168,
4367
+ "step": 681
4368
+ },
4369
+ {
4370
+ "epoch": 0.98,
4371
+ "learning_rate": 1.78897092465935e-06,
4372
+ "loss": 0.3628,
4373
+ "step": 682
4374
+ },
4375
+ {
4376
+ "epoch": 0.98,
4377
+ "learning_rate": 1.5849642338527926e-06,
4378
+ "loss": 0.6984,
4379
+ "step": 683
4380
+ },
4381
+ {
4382
+ "epoch": 0.98,
4383
+ "learning_rate": 1.3932581428751867e-06,
4384
+ "loss": 0.64,
4385
+ "step": 684
4386
+ },
4387
+ {
4388
+ "epoch": 0.98,
4389
+ "learning_rate": 1.2138645364101031e-06,
4390
+ "loss": 0.5111,
4391
+ "step": 685
4392
+ },
4393
+ {
4394
+ "epoch": 0.98,
4395
+ "learning_rate": 1.0467945358372921e-06,
4396
+ "loss": 0.6078,
4397
+ "step": 686
4398
+ },
4399
+ {
4400
+ "epoch": 0.98,
4401
+ "learning_rate": 8.920584985433023e-07,
4402
+ "loss": 0.368,
4403
+ "step": 687
4404
+ },
4405
+ {
4406
+ "epoch": 0.98,
4407
+ "learning_rate": 7.496660172793046e-07,
4408
+ "loss": 0.714,
4409
+ "step": 688
4410
+ },
4411
+ {
4412
+ "epoch": 0.99,
4413
+ "learning_rate": 6.196259195665244e-07,
4414
+ "loss": 0.3712,
4415
+ "step": 689
4416
+ },
4417
+ {
4418
+ "epoch": 0.99,
4419
+ "learning_rate": 5.019462671488562e-07,
4420
+ "loss": 0.5028,
4421
+ "step": 690
4422
+ },
4423
+ {
4424
+ "epoch": 0.99,
4425
+ "learning_rate": 3.9663435549308625e-07,
4426
+ "loss": 0.6019,
4427
+ "step": 691
4428
+ },
4429
+ {
4430
+ "epoch": 0.99,
4431
+ "learning_rate": 3.036967133367652e-07,
4432
+ "loss": 0.5255,
4433
+ "step": 692
4434
+ },
4435
+ {
4436
+ "epoch": 0.99,
4437
+ "learning_rate": 2.2313910228324297e-07,
4438
+ "loss": 0.5835,
4439
+ "step": 693
4440
+ },
4441
+ {
4442
+ "epoch": 0.99,
4443
+ "learning_rate": 1.5496651644468786e-07,
4444
+ "loss": 0.391,
4445
+ "step": 694
4446
+ },
4447
+ {
4448
+ "epoch": 0.99,
4449
+ "learning_rate": 9.918318213231193e-08,
4450
+ "loss": 0.486,
4451
+ "step": 695
4452
+ },
4453
+ {
4454
+ "epoch": 1.0,
4455
+ "learning_rate": 5.579255759451396e-08,
4456
+ "loss": 0.6019,
4457
+ "step": 696
4458
+ },
4459
+ {
4460
+ "epoch": 1.0,
4461
+ "learning_rate": 2.479733280236207e-08,
4462
+ "loss": 0.4126,
4463
+ "step": 697
4464
+ },
4465
+ {
4466
+ "epoch": 1.0,
4467
+ "learning_rate": 6.199429282904845e-09,
4468
+ "loss": 0.4922,
4469
+ "step": 698
4470
+ },
4471
+ {
4472
+ "epoch": 1.0,
4473
+ "learning_rate": 0.0,
4474
+ "loss": 0.4515,
4475
+ "step": 699
4476
+ }
4477
+ ],
4478
+ "logging_steps": 1,
4479
+ "max_steps": 699,
4480
+ "num_train_epochs": 1,
4481
+ "save_steps": 500,
4482
+ "total_flos": 1.8006457547803853e+18,
4483
+ "trial_name": null,
4484
+ "trial_params": null
4485
+ }
checkpoint-699/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1eafcb8aa53bf0edd51318ad38183fb1c8ab8c8b95a90eb8d6bd5abab0c2c463
3
+ size 4475
config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "meta-llama/Llama-2-13b-hf",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "bos_token_id": 1,
8
+ "eos_token_id": 2,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 5120,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 13824,
13
+ "max_position_embeddings": 4096,
14
+ "model_type": "llama",
15
+ "num_attention_heads": 40,
16
+ "num_hidden_layers": 40,
17
+ "num_key_value_heads": 40,
18
+ "pretraining_tp": 1,
19
+ "quantization_config": {
20
+ "bnb_4bit_compute_dtype": "bfloat16",
21
+ "bnb_4bit_quant_type": "nf4",
22
+ "bnb_4bit_use_double_quant": true,
23
+ "llm_int8_enable_fp32_cpu_offload": false,
24
+ "llm_int8_has_fp16_weight": false,
25
+ "llm_int8_skip_modules": null,
26
+ "llm_int8_threshold": 6.0,
27
+ "load_in_4bit": true,
28
+ "load_in_8bit": false,
29
+ "quant_method": "bitsandbytes"
30
+ },
31
+ "rms_norm_eps": 1e-05,
32
+ "rope_scaling": null,
33
+ "rope_theta": 10000.0,
34
+ "tie_word_embeddings": false,
35
+ "torch_dtype": "float16",
36
+ "transformers_version": "4.34.1",
37
+ "use_cache": false,
38
+ "vocab_size": 32000
39
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "</s>",
17
+ "unk_token": {
18
+ "content": "<unk>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
tokenizer_config.json ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "0": {
6
+ "content": "<unk>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "1": {
14
+ "content": "<s>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "2": {
22
+ "content": "</s>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ }
29
+ },
30
+ "bos_token": "<s>",
31
+ "clean_up_tokenization_spaces": false,
32
+ "eos_token": "</s>",
33
+ "legacy": false,
34
+ "model_max_length": 1000000000000000019884624838656,
35
+ "pad_token": "</s>",
36
+ "padding_side": "right",
37
+ "sp_model_kwargs": {},
38
+ "spaces_between_special_tokens": false,
39
+ "tokenizer_class": "LlamaTokenizer",
40
+ "trust_remote_code": false,
41
+ "unk_token": "<unk>",
42
+ "use_default_system_prompt": true,
43
+ "use_fast": true
44
+ }