Hugging Face's logo

Undi95
/
toxicqa-Llama2-13B-lora

PEFT
Safetensors
llama
Generated from Trainer
8-bit precision
bitsandbytes
Model card Files Community
Undi95 commited on
Commit
5f15ad4
·
1 Parent(s): 7a54a4d

Upload folder using huggingface_hub

Browse files
Files changed (23) hide show
  1. README.md +164 -0
  2. adapter_config.json +28 -0
  3. adapter_model.bin +3 -0
  4. checkpoint-1214/README.md +220 -0
  5. checkpoint-1214/adapter_config.json +28 -0
  6. checkpoint-1214/adapter_model.safetensors +3 -0
  7. checkpoint-1214/optimizer.pt +3 -0
  8. checkpoint-1214/rng_state.pth +3 -0
  9. checkpoint-1214/scheduler.pt +3 -0
  10. checkpoint-1214/trainer_state.json +0 -0
  11. checkpoint-1214/training_args.bin +3 -0
  12. checkpoint-607/README.md +220 -0
  13. checkpoint-607/adapter_config.json +28 -0
  14. checkpoint-607/adapter_model.safetensors +3 -0
  15. checkpoint-607/optimizer.pt +3 -0
  16. checkpoint-607/rng_state.pth +3 -0
  17. checkpoint-607/scheduler.pt +3 -0
  18. checkpoint-607/trainer_state.json +3695 -0
  19. checkpoint-607/training_args.bin +3 -0
  20. config.json +41 -0
  21. special_tokens_map.json +30 -0
  22. tokenizer.model +3 -0
  23. tokenizer_config.json +43 -0
README.md ADDED
@@ -0,0 +1,164 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ tags:
4
+ - generated_from_trainer
5
+ base_model: NousResearch/Llama-2-13b-hf
6
+ model-index:
7
+ - name: lora-out
8
+ results: []
9
+ ---
10
+
11
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
12
+ should probably proofread and complete it, then remove this comment. -->
13
+
14
+ [<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)
15
+ <details><summary>See axolotl config</summary>
16
+
17
+ axolotl version: `0.3.0`
18
+ ```yaml
19
+ base_model: NousResearch/Llama-2-13b-hf
20
+ model_type: LlamaForCausalLM
21
+ tokenizer_type: LlamaTokenizer
22
+ is_llama_derived_model: true
23
+
24
+ load_in_8bit: true
25
+ load_in_4bit: false
26
+ strict: false
27
+
28
+ datasets:
29
+ - path: dataset
30
+ type: sharegpt
31
+ dataset_prepared_path:
32
+ val_set_size: 0.05
33
+ output_dir: ./lora-out
34
+
35
+ sequence_len: 4096
36
+ sample_packing: true
37
+ pad_to_sequence_len: true
38
+
39
+ adapter: lora
40
+ lora_model_dir:
41
+ lora_r: 128
42
+ lora_alpha: 64
43
+ lora_dropout: 0.05
44
+ lora_target_linear: true
45
+ lora_fan_in_fan_out:
46
+
47
+ wandb_project: toxicLlama-2-13B
48
+ wandb_entity:
49
+ wandb_watch:
50
+ wandb_name:
51
+ wandb_log_model:
52
+
53
+ gradient_accumulation_steps: 1
54
+ micro_batch_size: 2
55
+ num_epochs: 2
56
+ optimizer: adamw_bnb_8bit
57
+ lr_scheduler: cosine
58
+ learning_rate: 0.0002
59
+ eval_batch_size: 2
60
+
61
+ train_on_inputs: false
62
+ group_by_length: false
63
+ bf16: true
64
+ fp16: false
65
+ tf32: false
66
+
67
+ gradient_checkpointing: true
68
+ early_stopping_patience:
69
+ resume_from_checkpoint:
70
+ local_rank:
71
+ logging_steps: 1
72
+ xformers_attention:
73
+ flash_attention: true
74
+
75
+ warmup_steps: 10
76
+ evals_per_epoch: 4
77
+ eval_table_size:
78
+ eval_table_max_new_tokens: 128
79
+ saves_per_epoch: 1
80
+ debug:
81
+ deepspeed:
82
+ weight_decay: 0.0
83
+ fsdp:
84
+ fsdp_config:
85
+ special_tokens:
86
+ bos_token: "<s>"
87
+ eos_token: "</s>"
88
+ unk_token: "<unk>"
89
+
90
+ ```
91
+
92
+ </details><br>
93
+
94
+ # lora-out
95
+
96
+ This model is a fine-tuned version of [NousResearch/Llama-2-13b-hf](https://huggingface.co/NousResearch/Llama-2-13b-hf) on the None dataset.
97
+ It achieves the following results on the evaluation set:
98
+ - Loss: 0.7634
99
+
100
+ ## Model description
101
+
102
+ More information needed
103
+
104
+ ## Intended uses & limitations
105
+
106
+ More information needed
107
+
108
+ ## Training and evaluation data
109
+
110
+ More information needed
111
+
112
+ ## Training procedure
113
+
114
+ ### Training hyperparameters
115
+
116
+ The following hyperparameters were used during training:
117
+ - learning_rate: 0.0002
118
+ - train_batch_size: 2
119
+ - eval_batch_size: 2
120
+ - seed: 42
121
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
122
+ - lr_scheduler_type: cosine
123
+ - lr_scheduler_warmup_steps: 10
124
+ - num_epochs: 2
125
+
126
+ ### Training results
127
+
128
+ | Training Loss | Epoch | Step | Validation Loss |
129
+ |:-------------:|:-----:|:----:|:---------------:|
130
+ | 1.0107 | 0.0 | 1 | 1.0286 |
131
+ | 0.8198 | 0.25 | 152 | 0.8079 |
132
+ | 0.7993 | 0.5 | 304 | 0.7904 |
133
+ | 0.7348 | 0.75 | 456 | 0.7748 |
134
+ | 0.689 | 1.0 | 608 | 0.7638 |
135
+ | 0.6462 | 1.23 | 760 | 0.7729 |
136
+ | 0.6226 | 1.48 | 912 | 0.7657 |
137
+ | 0.6179 | 1.73 | 1064 | 0.7634 |
138
+
139
+
140
+ ### Framework versions
141
+
142
+ - Transformers 4.36.2
143
+ - Pytorch 2.0.1+cu118
144
+ - Datasets 2.16.1
145
+ - Tokenizers 0.15.0
146
+ ## Training procedure
147
+
148
+
149
+ The following `bitsandbytes` quantization config was used during training:
150
+ - quant_method: bitsandbytes
151
+ - load_in_8bit: True
152
+ - load_in_4bit: False
153
+ - llm_int8_threshold: 6.0
154
+ - llm_int8_skip_modules: None
155
+ - llm_int8_enable_fp32_cpu_offload: False
156
+ - llm_int8_has_fp16_weight: False
157
+ - bnb_4bit_quant_type: fp4
158
+ - bnb_4bit_use_double_quant: False
159
+ - bnb_4bit_compute_dtype: float32
160
+
161
+ ### Framework versions
162
+
163
+
164
+ - PEFT 0.6.0
adapter_config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "NousResearch/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": 128,
16
+ "rank_pattern": {},
17
+ "revision": null,
18
+ "target_modules": [
19
+ "k_proj",
20
+ "gate_proj",
21
+ "up_proj",
22
+ "v_proj",
23
+ "o_proj",
24
+ "q_proj",
25
+ "down_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:c6972544537069f45f6a8bf8768ce08f1bfd8f37220c7bad0b453729ab90479c
3
+ size 2002982221
checkpoint-1214/README.md ADDED
@@ -0,0 +1,220 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ base_model: NousResearch/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
+ - **Funded by [optional]:** [More Information Needed]
22
+ - **Shared by [optional]:** [More Information Needed]
23
+ - **Model type:** [More Information Needed]
24
+ - **Language(s) (NLP):** [More Information Needed]
25
+ - **License:** [More Information Needed]
26
+ - **Finetuned from model [optional]:** [More Information Needed]
27
+
28
+ ### Model Sources [optional]
29
+
30
+ <!-- Provide the basic links for the model. -->
31
+
32
+ - **Repository:** [More Information Needed]
33
+ - **Paper [optional]:** [More Information Needed]
34
+ - **Demo [optional]:** [More Information Needed]
35
+
36
+ ## Uses
37
+
38
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
39
+
40
+ ### Direct Use
41
+
42
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
43
+
44
+ [More Information Needed]
45
+
46
+ ### Downstream Use [optional]
47
+
48
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
49
+
50
+ [More Information Needed]
51
+
52
+ ### Out-of-Scope Use
53
+
54
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
55
+
56
+ [More Information Needed]
57
+
58
+ ## Bias, Risks, and Limitations
59
+
60
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
61
+
62
+ [More Information Needed]
63
+
64
+ ### Recommendations
65
+
66
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
67
+
68
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
69
+
70
+ ## How to Get Started with the Model
71
+
72
+ Use the code below to get started with the model.
73
+
74
+ [More Information Needed]
75
+
76
+ ## Training Details
77
+
78
+ ### Training Data
79
+
80
+ <!-- This should link to a Dataset 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. -->
81
+
82
+ [More Information Needed]
83
+
84
+ ### Training Procedure
85
+
86
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
87
+
88
+ #### Preprocessing [optional]
89
+
90
+ [More Information Needed]
91
+
92
+
93
+ #### Training Hyperparameters
94
+
95
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
96
+
97
+ #### Speeds, Sizes, Times [optional]
98
+
99
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
100
+
101
+ [More Information Needed]
102
+
103
+ ## Evaluation
104
+
105
+ <!-- This section describes the evaluation protocols and provides the results. -->
106
+
107
+ ### Testing Data, Factors & Metrics
108
+
109
+ #### Testing Data
110
+
111
+ <!-- This should link to a Dataset Card if possible. -->
112
+
113
+ [More Information Needed]
114
+
115
+ #### Factors
116
+
117
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
118
+
119
+ [More Information Needed]
120
+
121
+ #### Metrics
122
+
123
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
124
+
125
+ [More Information Needed]
126
+
127
+ ### Results
128
+
129
+ [More Information Needed]
130
+
131
+ #### Summary
132
+
133
+
134
+
135
+ ## Model Examination [optional]
136
+
137
+ <!-- Relevant interpretability work for the model goes here -->
138
+
139
+ [More Information Needed]
140
+
141
+ ## Environmental Impact
142
+
143
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
144
+
145
+ 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).
146
+
147
+ - **Hardware Type:** [More Information Needed]
148
+ - **Hours used:** [More Information Needed]
149
+ - **Cloud Provider:** [More Information Needed]
150
+ - **Compute Region:** [More Information Needed]
151
+ - **Carbon Emitted:** [More Information Needed]
152
+
153
+ ## Technical Specifications [optional]
154
+
155
+ ### Model Architecture and Objective
156
+
157
+ [More Information Needed]
158
+
159
+ ### Compute Infrastructure
160
+
161
+ [More Information Needed]
162
+
163
+ #### Hardware
164
+
165
+ [More Information Needed]
166
+
167
+ #### Software
168
+
169
+ [More Information Needed]
170
+
171
+ ## Citation [optional]
172
+
173
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
174
+
175
+ **BibTeX:**
176
+
177
+ [More Information Needed]
178
+
179
+ **APA:**
180
+
181
+ [More Information Needed]
182
+
183
+ ## Glossary [optional]
184
+
185
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
186
+
187
+ [More Information Needed]
188
+
189
+ ## More Information [optional]
190
+
191
+ [More Information Needed]
192
+
193
+ ## Model Card Authors [optional]
194
+
195
+ [More Information Needed]
196
+
197
+ ## Model Card Contact
198
+
199
+ [More Information Needed]
200
+
201
+
202
+ ## Training procedure
203
+
204
+
205
+ The following `bitsandbytes` quantization config was used during training:
206
+ - quant_method: bitsandbytes
207
+ - load_in_8bit: True
208
+ - load_in_4bit: False
209
+ - llm_int8_threshold: 6.0
210
+ - llm_int8_skip_modules: None
211
+ - llm_int8_enable_fp32_cpu_offload: False
212
+ - llm_int8_has_fp16_weight: False
213
+ - bnb_4bit_quant_type: fp4
214
+ - bnb_4bit_use_double_quant: False
215
+ - bnb_4bit_compute_dtype: float32
216
+
217
+ ### Framework versions
218
+
219
+
220
+ - PEFT 0.6.0
checkpoint-1214/adapter_config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "NousResearch/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": 128,
16
+ "rank_pattern": {},
17
+ "revision": null,
18
+ "target_modules": [
19
+ "k_proj",
20
+ "gate_proj",
21
+ "up_proj",
22
+ "v_proj",
23
+ "o_proj",
24
+ "q_proj",
25
+ "down_proj"
26
+ ],
27
+ "task_type": "CAUSAL_LM"
28
+ }
checkpoint-1214/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17e509881b2ce3171040cef64154301cda4579863cf1779789c802fbaabe1738
3
+ size 2002857080
checkpoint-1214/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:76d15c601313d80e9b80af2eb9b989de953cad19ae068350593dd1430acf949d
3
+ size 1004004575
checkpoint-1214/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1a4c282d969f2f4f599d9a1949cd2d82c0510cacd751917f558ef5ca9164d8f
3
+ size 14575
checkpoint-1214/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:27d0c53fea9fb85336a27a8a5849d435c9daa0e9eddbae1615ab38d922349479
3
+ size 627
checkpoint-1214/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1214/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0b7fb8249d23a78d379d117f7117168f666a5d46ad4a572405b96de4612aa1d7
3
+ size 4731
checkpoint-607/README.md ADDED
@@ -0,0 +1,220 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ base_model: NousResearch/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
+ - **Funded by [optional]:** [More Information Needed]
22
+ - **Shared by [optional]:** [More Information Needed]
23
+ - **Model type:** [More Information Needed]
24
+ - **Language(s) (NLP):** [More Information Needed]
25
+ - **License:** [More Information Needed]
26
+ - **Finetuned from model [optional]:** [More Information Needed]
27
+
28
+ ### Model Sources [optional]
29
+
30
+ <!-- Provide the basic links for the model. -->
31
+
32
+ - **Repository:** [More Information Needed]
33
+ - **Paper [optional]:** [More Information Needed]
34
+ - **Demo [optional]:** [More Information Needed]
35
+
36
+ ## Uses
37
+
38
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
39
+
40
+ ### Direct Use
41
+
42
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
43
+
44
+ [More Information Needed]
45
+
46
+ ### Downstream Use [optional]
47
+
48
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
49
+
50
+ [More Information Needed]
51
+
52
+ ### Out-of-Scope Use
53
+
54
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
55
+
56
+ [More Information Needed]
57
+
58
+ ## Bias, Risks, and Limitations
59
+
60
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
61
+
62
+ [More Information Needed]
63
+
64
+ ### Recommendations
65
+
66
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
67
+
68
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
69
+
70
+ ## How to Get Started with the Model
71
+
72
+ Use the code below to get started with the model.
73
+
74
+ [More Information Needed]
75
+
76
+ ## Training Details
77
+
78
+ ### Training Data
79
+
80
+ <!-- This should link to a Dataset 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. -->
81
+
82
+ [More Information Needed]
83
+
84
+ ### Training Procedure
85
+
86
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
87
+
88
+ #### Preprocessing [optional]
89
+
90
+ [More Information Needed]
91
+
92
+
93
+ #### Training Hyperparameters
94
+
95
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
96
+
97
+ #### Speeds, Sizes, Times [optional]
98
+
99
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
100
+
101
+ [More Information Needed]
102
+
103
+ ## Evaluation
104
+
105
+ <!-- This section describes the evaluation protocols and provides the results. -->
106
+
107
+ ### Testing Data, Factors & Metrics
108
+
109
+ #### Testing Data
110
+
111
+ <!-- This should link to a Dataset Card if possible. -->
112
+
113
+ [More Information Needed]
114
+
115
+ #### Factors
116
+
117
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
118
+
119
+ [More Information Needed]
120
+
121
+ #### Metrics
122
+
123
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
124
+
125
+ [More Information Needed]
126
+
127
+ ### Results
128
+
129
+ [More Information Needed]
130
+
131
+ #### Summary
132
+
133
+
134
+
135
+ ## Model Examination [optional]
136
+
137
+ <!-- Relevant interpretability work for the model goes here -->
138
+
139
+ [More Information Needed]
140
+
141
+ ## Environmental Impact
142
+
143
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
144
+
145
+ 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).
146
+
147
+ - **Hardware Type:** [More Information Needed]
148
+ - **Hours used:** [More Information Needed]
149
+ - **Cloud Provider:** [More Information Needed]
150
+ - **Compute Region:** [More Information Needed]
151
+ - **Carbon Emitted:** [More Information Needed]
152
+
153
+ ## Technical Specifications [optional]
154
+
155
+ ### Model Architecture and Objective
156
+
157
+ [More Information Needed]
158
+
159
+ ### Compute Infrastructure
160
+
161
+ [More Information Needed]
162
+
163
+ #### Hardware
164
+
165
+ [More Information Needed]
166
+
167
+ #### Software
168
+
169
+ [More Information Needed]
170
+
171
+ ## Citation [optional]
172
+
173
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
174
+
175
+ **BibTeX:**
176
+
177
+ [More Information Needed]
178
+
179
+ **APA:**
180
+
181
+ [More Information Needed]
182
+
183
+ ## Glossary [optional]
184
+
185
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
186
+
187
+ [More Information Needed]
188
+
189
+ ## More Information [optional]
190
+
191
+ [More Information Needed]
192
+
193
+ ## Model Card Authors [optional]
194
+
195
+ [More Information Needed]
196
+
197
+ ## Model Card Contact
198
+
199
+ [More Information Needed]
200
+
201
+
202
+ ## Training procedure
203
+
204
+
205
+ The following `bitsandbytes` quantization config was used during training:
206
+ - quant_method: bitsandbytes
207
+ - load_in_8bit: True
208
+ - load_in_4bit: False
209
+ - llm_int8_threshold: 6.0
210
+ - llm_int8_skip_modules: None
211
+ - llm_int8_enable_fp32_cpu_offload: False
212
+ - llm_int8_has_fp16_weight: False
213
+ - bnb_4bit_quant_type: fp4
214
+ - bnb_4bit_use_double_quant: False
215
+ - bnb_4bit_compute_dtype: float32
216
+
217
+ ### Framework versions
218
+
219
+
220
+ - PEFT 0.6.0
checkpoint-607/adapter_config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "NousResearch/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": 128,
16
+ "rank_pattern": {},
17
+ "revision": null,
18
+ "target_modules": [
19
+ "k_proj",
20
+ "gate_proj",
21
+ "up_proj",
22
+ "v_proj",
23
+ "o_proj",
24
+ "q_proj",
25
+ "down_proj"
26
+ ],
27
+ "task_type": "CAUSAL_LM"
28
+ }
checkpoint-607/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dc4d64f092e63ff0b9ad029f73f257008fd916338fda71e4bf0d823aefbb4026
3
+ size 2002857080
checkpoint-607/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2233a4829408725752c08af7ff92948c8008d53591318da2692ca8a759781e4c
3
+ size 1004004575
checkpoint-607/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:067b17c920e82786b2f9e4525b48f8fc5019fe733e45e4bf9f6bfc963b38786b
3
+ size 14575
checkpoint-607/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5f35da3ff1f52174931d7db7337984711f946c9e67482bf30cde8fe03857e6d8
3
+ size 627
checkpoint-607/trainer_state.json ADDED
@@ -0,0 +1,3695 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "eval_steps": 152,
6
+ "global_step": 607,
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": 2e-05,
14
+ "loss": 1.0107,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "eval_loss": 1.0285824537277222,
20
+ "eval_runtime": 85.1275,
21
+ "eval_samples_per_second": 4.041,
22
+ "eval_steps_per_second": 2.02,
23
+ "step": 1
24
+ },
25
+ {
26
+ "epoch": 0.0,
27
+ "learning_rate": 4e-05,
28
+ "loss": 1.0286,
29
+ "step": 2
30
+ },
31
+ {
32
+ "epoch": 0.0,
33
+ "learning_rate": 6e-05,
34
+ "loss": 1.0509,
35
+ "step": 3
36
+ },
37
+ {
38
+ "epoch": 0.01,
39
+ "learning_rate": 8e-05,
40
+ "loss": 1.0267,
41
+ "step": 4
42
+ },
43
+ {
44
+ "epoch": 0.01,
45
+ "learning_rate": 0.0001,
46
+ "loss": 1.0035,
47
+ "step": 5
48
+ },
49
+ {
50
+ "epoch": 0.01,
51
+ "learning_rate": 0.00012,
52
+ "loss": 0.98,
53
+ "step": 6
54
+ },
55
+ {
56
+ "epoch": 0.01,
57
+ "learning_rate": 0.00014,
58
+ "loss": 0.8973,
59
+ "step": 7
60
+ },
61
+ {
62
+ "epoch": 0.01,
63
+ "learning_rate": 0.00016,
64
+ "loss": 0.9456,
65
+ "step": 8
66
+ },
67
+ {
68
+ "epoch": 0.01,
69
+ "learning_rate": 0.00018,
70
+ "loss": 0.9843,
71
+ "step": 9
72
+ },
73
+ {
74
+ "epoch": 0.02,
75
+ "learning_rate": 0.0002,
76
+ "loss": 0.926,
77
+ "step": 10
78
+ },
79
+ {
80
+ "epoch": 0.02,
81
+ "learning_rate": 0.00019999965957885393,
82
+ "loss": 0.9399,
83
+ "step": 11
84
+ },
85
+ {
86
+ "epoch": 0.02,
87
+ "learning_rate": 0.0001999986383177335,
88
+ "loss": 0.8926,
89
+ "step": 12
90
+ },
91
+ {
92
+ "epoch": 0.02,
93
+ "learning_rate": 0.00019999693622359184,
94
+ "loss": 0.9118,
95
+ "step": 13
96
+ },
97
+ {
98
+ "epoch": 0.02,
99
+ "learning_rate": 0.0001999945533080175,
100
+ "loss": 0.9197,
101
+ "step": 14
102
+ },
103
+ {
104
+ "epoch": 0.02,
105
+ "learning_rate": 0.00019999148958723447,
106
+ "loss": 0.9105,
107
+ "step": 15
108
+ },
109
+ {
110
+ "epoch": 0.03,
111
+ "learning_rate": 0.0001999877450821018,
112
+ "loss": 0.8527,
113
+ "step": 16
114
+ },
115
+ {
116
+ "epoch": 0.03,
117
+ "learning_rate": 0.00019998331981811366,
118
+ "loss": 0.8422,
119
+ "step": 17
120
+ },
121
+ {
122
+ "epoch": 0.03,
123
+ "learning_rate": 0.0001999782138253991,
124
+ "loss": 0.8741,
125
+ "step": 18
126
+ },
127
+ {
128
+ "epoch": 0.03,
129
+ "learning_rate": 0.00019997242713872196,
130
+ "loss": 0.8792,
131
+ "step": 19
132
+ },
133
+ {
134
+ "epoch": 0.03,
135
+ "learning_rate": 0.00019996595979748037,
136
+ "loss": 0.8766,
137
+ "step": 20
138
+ },
139
+ {
140
+ "epoch": 0.03,
141
+ "learning_rate": 0.00019995881184570676,
142
+ "loss": 0.8361,
143
+ "step": 21
144
+ },
145
+ {
146
+ "epoch": 0.04,
147
+ "learning_rate": 0.00019995098333206742,
148
+ "loss": 0.8709,
149
+ "step": 22
150
+ },
151
+ {
152
+ "epoch": 0.04,
153
+ "learning_rate": 0.00019994247430986213,
154
+ "loss": 0.8575,
155
+ "step": 23
156
+ },
157
+ {
158
+ "epoch": 0.04,
159
+ "learning_rate": 0.00019993328483702393,
160
+ "loss": 0.8293,
161
+ "step": 24
162
+ },
163
+ {
164
+ "epoch": 0.04,
165
+ "learning_rate": 0.0001999234149761187,
166
+ "loss": 0.8589,
167
+ "step": 25
168
+ },
169
+ {
170
+ "epoch": 0.04,
171
+ "learning_rate": 0.00019991286479434454,
172
+ "loss": 0.8203,
173
+ "step": 26
174
+ },
175
+ {
176
+ "epoch": 0.04,
177
+ "learning_rate": 0.0001999016343635316,
178
+ "loss": 0.8196,
179
+ "step": 27
180
+ },
181
+ {
182
+ "epoch": 0.05,
183
+ "learning_rate": 0.00019988972376014142,
184
+ "loss": 0.8936,
185
+ "step": 28
186
+ },
187
+ {
188
+ "epoch": 0.05,
189
+ "learning_rate": 0.00019987713306526638,
190
+ "loss": 0.9094,
191
+ "step": 29
192
+ },
193
+ {
194
+ "epoch": 0.05,
195
+ "learning_rate": 0.00019986386236462924,
196
+ "loss": 0.8147,
197
+ "step": 30
198
+ },
199
+ {
200
+ "epoch": 0.05,
201
+ "learning_rate": 0.00019984991174858257,
202
+ "loss": 0.875,
203
+ "step": 31
204
+ },
205
+ {
206
+ "epoch": 0.05,
207
+ "learning_rate": 0.00019983528131210812,
208
+ "loss": 0.8109,
209
+ "step": 32
210
+ },
211
+ {
212
+ "epoch": 0.05,
213
+ "learning_rate": 0.00019981997115481602,
214
+ "loss": 0.8138,
215
+ "step": 33
216
+ },
217
+ {
218
+ "epoch": 0.06,
219
+ "learning_rate": 0.00019980398138094428,
220
+ "loss": 0.786,
221
+ "step": 34
222
+ },
223
+ {
224
+ "epoch": 0.06,
225
+ "learning_rate": 0.0001997873120993581,
226
+ "loss": 0.8516,
227
+ "step": 35
228
+ },
229
+ {
230
+ "epoch": 0.06,
231
+ "learning_rate": 0.00019976996342354898,
232
+ "loss": 0.8395,
233
+ "step": 36
234
+ },
235
+ {
236
+ "epoch": 0.06,
237
+ "learning_rate": 0.00019975193547163404,
238
+ "loss": 0.8347,
239
+ "step": 37
240
+ },
241
+ {
242
+ "epoch": 0.06,
243
+ "learning_rate": 0.00019973322836635518,
244
+ "loss": 0.8074,
245
+ "step": 38
246
+ },
247
+ {
248
+ "epoch": 0.06,
249
+ "learning_rate": 0.0001997138422350783,
250
+ "loss": 0.8411,
251
+ "step": 39
252
+ },
253
+ {
254
+ "epoch": 0.07,
255
+ "learning_rate": 0.00019969377720979237,
256
+ "loss": 0.8197,
257
+ "step": 40
258
+ },
259
+ {
260
+ "epoch": 0.07,
261
+ "learning_rate": 0.00019967303342710864,
262
+ "loss": 0.8325,
263
+ "step": 41
264
+ },
265
+ {
266
+ "epoch": 0.07,
267
+ "learning_rate": 0.00019965161102825945,
268
+ "loss": 0.8291,
269
+ "step": 42
270
+ },
271
+ {
272
+ "epoch": 0.07,
273
+ "learning_rate": 0.0001996295101590976,
274
+ "loss": 0.8414,
275
+ "step": 43
276
+ },
277
+ {
278
+ "epoch": 0.07,
279
+ "learning_rate": 0.00019960673097009518,
280
+ "loss": 0.85,
281
+ "step": 44
282
+ },
283
+ {
284
+ "epoch": 0.07,
285
+ "learning_rate": 0.00019958327361634248,
286
+ "loss": 0.8324,
287
+ "step": 45
288
+ },
289
+ {
290
+ "epoch": 0.08,
291
+ "learning_rate": 0.00019955913825754713,
292
+ "loss": 0.8225,
293
+ "step": 46
294
+ },
295
+ {
296
+ "epoch": 0.08,
297
+ "learning_rate": 0.00019953432505803286,
298
+ "loss": 0.8364,
299
+ "step": 47
300
+ },
301
+ {
302
+ "epoch": 0.08,
303
+ "learning_rate": 0.0001995088341867384,
304
+ "loss": 0.8513,
305
+ "step": 48
306
+ },
307
+ {
308
+ "epoch": 0.08,
309
+ "learning_rate": 0.00019948266581721642,
310
+ "loss": 0.8739,
311
+ "step": 49
312
+ },
313
+ {
314
+ "epoch": 0.08,
315
+ "learning_rate": 0.0001994558201276322,
316
+ "loss": 0.8212,
317
+ "step": 50
318
+ },
319
+ {
320
+ "epoch": 0.08,
321
+ "learning_rate": 0.00019942829730076257,
322
+ "loss": 0.8821,
323
+ "step": 51
324
+ },
325
+ {
326
+ "epoch": 0.09,
327
+ "learning_rate": 0.0001994000975239946,
328
+ "loss": 0.8511,
329
+ "step": 52
330
+ },
331
+ {
332
+ "epoch": 0.09,
333
+ "learning_rate": 0.00019937122098932428,
334
+ "loss": 0.8369,
335
+ "step": 53
336
+ },
337
+ {
338
+ "epoch": 0.09,
339
+ "learning_rate": 0.00019934166789335525,
340
+ "loss": 0.8224,
341
+ "step": 54
342
+ },
343
+ {
344
+ "epoch": 0.09,
345
+ "learning_rate": 0.00019931143843729748,
346
+ "loss": 0.7896,
347
+ "step": 55
348
+ },
349
+ {
350
+ "epoch": 0.09,
351
+ "learning_rate": 0.00019928053282696596,
352
+ "loss": 0.8295,
353
+ "step": 56
354
+ },
355
+ {
356
+ "epoch": 0.09,
357
+ "learning_rate": 0.00019924895127277907,
358
+ "loss": 0.8231,
359
+ "step": 57
360
+ },
361
+ {
362
+ "epoch": 0.1,
363
+ "learning_rate": 0.00019921669398975745,
364
+ "loss": 0.8321,
365
+ "step": 58
366
+ },
367
+ {
368
+ "epoch": 0.1,
369
+ "learning_rate": 0.0001991837611975223,
370
+ "loss": 0.8051,
371
+ "step": 59
372
+ },
373
+ {
374
+ "epoch": 0.1,
375
+ "learning_rate": 0.000199150153120294,
376
+ "loss": 0.8772,
377
+ "step": 60
378
+ },
379
+ {
380
+ "epoch": 0.1,
381
+ "learning_rate": 0.0001991158699868905,
382
+ "loss": 0.8126,
383
+ "step": 61
384
+ },
385
+ {
386
+ "epoch": 0.1,
387
+ "learning_rate": 0.00019908091203072598,
388
+ "loss": 0.8615,
389
+ "step": 62
390
+ },
391
+ {
392
+ "epoch": 0.1,
393
+ "learning_rate": 0.00019904527948980894,
394
+ "loss": 0.8272,
395
+ "step": 63
396
+ },
397
+ {
398
+ "epoch": 0.11,
399
+ "learning_rate": 0.00019900897260674073,
400
+ "loss": 0.8134,
401
+ "step": 64
402
+ },
403
+ {
404
+ "epoch": 0.11,
405
+ "learning_rate": 0.00019897199162871408,
406
+ "loss": 0.8558,
407
+ "step": 65
408
+ },
409
+ {
410
+ "epoch": 0.11,
411
+ "learning_rate": 0.00019893433680751103,
412
+ "loss": 0.8155,
413
+ "step": 66
414
+ },
415
+ {
416
+ "epoch": 0.11,
417
+ "learning_rate": 0.00019889600839950155,
418
+ "loss": 0.8504,
419
+ "step": 67
420
+ },
421
+ {
422
+ "epoch": 0.11,
423
+ "learning_rate": 0.0001988570066656417,
424
+ "loss": 0.7906,
425
+ "step": 68
426
+ },
427
+ {
428
+ "epoch": 0.11,
429
+ "learning_rate": 0.00019881733187147171,
430
+ "loss": 0.8246,
431
+ "step": 69
432
+ },
433
+ {
434
+ "epoch": 0.12,
435
+ "learning_rate": 0.00019877698428711442,
436
+ "loss": 0.8345,
437
+ "step": 70
438
+ },
439
+ {
440
+ "epoch": 0.12,
441
+ "learning_rate": 0.0001987359641872732,
442
+ "loss": 0.8161,
443
+ "step": 71
444
+ },
445
+ {
446
+ "epoch": 0.12,
447
+ "learning_rate": 0.00019869427185123027,
448
+ "loss": 0.8307,
449
+ "step": 72
450
+ },
451
+ {
452
+ "epoch": 0.12,
453
+ "learning_rate": 0.00019865190756284467,
454
+ "loss": 0.8157,
455
+ "step": 73
456
+ },
457
+ {
458
+ "epoch": 0.12,
459
+ "learning_rate": 0.00019860887161055038,
460
+ "loss": 0.8138,
461
+ "step": 74
462
+ },
463
+ {
464
+ "epoch": 0.12,
465
+ "learning_rate": 0.0001985651642873544,
466
+ "loss": 0.8162,
467
+ "step": 75
468
+ },
469
+ {
470
+ "epoch": 0.13,
471
+ "learning_rate": 0.00019852078589083466,
472
+ "loss": 0.8313,
473
+ "step": 76
474
+ },
475
+ {
476
+ "epoch": 0.13,
477
+ "learning_rate": 0.00019847573672313802,
478
+ "loss": 0.8081,
479
+ "step": 77
480
+ },
481
+ {
482
+ "epoch": 0.13,
483
+ "learning_rate": 0.0001984300170909783,
484
+ "loss": 0.7811,
485
+ "step": 78
486
+ },
487
+ {
488
+ "epoch": 0.13,
489
+ "learning_rate": 0.00019838362730563406,
490
+ "loss": 0.833,
491
+ "step": 79
492
+ },
493
+ {
494
+ "epoch": 0.13,
495
+ "learning_rate": 0.00019833656768294662,
496
+ "loss": 0.8214,
497
+ "step": 80
498
+ },
499
+ {
500
+ "epoch": 0.13,
501
+ "learning_rate": 0.00019828883854331776,
502
+ "loss": 0.8096,
503
+ "step": 81
504
+ },
505
+ {
506
+ "epoch": 0.14,
507
+ "learning_rate": 0.0001982404402117077,
508
+ "loss": 0.855,
509
+ "step": 82
510
+ },
511
+ {
512
+ "epoch": 0.14,
513
+ "learning_rate": 0.00019819137301763267,
514
+ "loss": 0.8122,
515
+ "step": 83
516
+ },
517
+ {
518
+ "epoch": 0.14,
519
+ "learning_rate": 0.00019814163729516292,
520
+ "loss": 0.8423,
521
+ "step": 84
522
+ },
523
+ {
524
+ "epoch": 0.14,
525
+ "learning_rate": 0.0001980912333829203,
526
+ "loss": 0.83,
527
+ "step": 85
528
+ },
529
+ {
530
+ "epoch": 0.14,
531
+ "learning_rate": 0.0001980401616240759,
532
+ "loss": 0.8439,
533
+ "step": 86
534
+ },
535
+ {
536
+ "epoch": 0.14,
537
+ "learning_rate": 0.00019798842236634797,
538
+ "loss": 0.7693,
539
+ "step": 87
540
+ },
541
+ {
542
+ "epoch": 0.14,
543
+ "learning_rate": 0.00019793601596199912,
544
+ "loss": 0.8443,
545
+ "step": 88
546
+ },
547
+ {
548
+ "epoch": 0.15,
549
+ "learning_rate": 0.00019788294276783442,
550
+ "loss": 0.8203,
551
+ "step": 89
552
+ },
553
+ {
554
+ "epoch": 0.15,
555
+ "learning_rate": 0.00019782920314519856,
556
+ "loss": 0.7918,
557
+ "step": 90
558
+ },
559
+ {
560
+ "epoch": 0.15,
561
+ "learning_rate": 0.00019777479745997366,
562
+ "loss": 0.856,
563
+ "step": 91
564
+ },
565
+ {
566
+ "epoch": 0.15,
567
+ "learning_rate": 0.00019771972608257659,
568
+ "loss": 0.8001,
569
+ "step": 92
570
+ },
571
+ {
572
+ "epoch": 0.15,
573
+ "learning_rate": 0.00019766398938795662,
574
+ "loss": 0.7688,
575
+ "step": 93
576
+ },
577
+ {
578
+ "epoch": 0.15,
579
+ "learning_rate": 0.00019760758775559274,
580
+ "loss": 0.8188,
581
+ "step": 94
582
+ },
583
+ {
584
+ "epoch": 0.16,
585
+ "learning_rate": 0.00019755052156949105,
586
+ "loss": 0.81,
587
+ "step": 95
588
+ },
589
+ {
590
+ "epoch": 0.16,
591
+ "learning_rate": 0.00019749279121818235,
592
+ "loss": 0.8473,
593
+ "step": 96
594
+ },
595
+ {
596
+ "epoch": 0.16,
597
+ "learning_rate": 0.0001974343970947193,
598
+ "loss": 0.7996,
599
+ "step": 97
600
+ },
601
+ {
602
+ "epoch": 0.16,
603
+ "learning_rate": 0.0001973753395966737,
604
+ "loss": 0.7884,
605
+ "step": 98
606
+ },
607
+ {
608
+ "epoch": 0.16,
609
+ "learning_rate": 0.00019731561912613406,
610
+ "loss": 0.829,
611
+ "step": 99
612
+ },
613
+ {
614
+ "epoch": 0.16,
615
+ "learning_rate": 0.00019725523608970255,
616
+ "loss": 0.7999,
617
+ "step": 100
618
+ },
619
+ {
620
+ "epoch": 0.17,
621
+ "learning_rate": 0.00019719419089849247,
622
+ "loss": 0.8344,
623
+ "step": 101
624
+ },
625
+ {
626
+ "epoch": 0.17,
627
+ "learning_rate": 0.00019713248396812524,
628
+ "loss": 0.8108,
629
+ "step": 102
630
+ },
631
+ {
632
+ "epoch": 0.17,
633
+ "learning_rate": 0.00019707011571872777,
634
+ "loss": 0.8543,
635
+ "step": 103
636
+ },
637
+ {
638
+ "epoch": 0.17,
639
+ "learning_rate": 0.00019700708657492948,
640
+ "loss": 0.7791,
641
+ "step": 104
642
+ },
643
+ {
644
+ "epoch": 0.17,
645
+ "learning_rate": 0.00019694339696585942,
646
+ "loss": 0.8245,
647
+ "step": 105
648
+ },
649
+ {
650
+ "epoch": 0.17,
651
+ "learning_rate": 0.0001968790473251434,
652
+ "loss": 0.8192,
653
+ "step": 106
654
+ },
655
+ {
656
+ "epoch": 0.18,
657
+ "learning_rate": 0.00019681403809090097,
658
+ "loss": 0.826,
659
+ "step": 107
660
+ },
661
+ {
662
+ "epoch": 0.18,
663
+ "learning_rate": 0.00019674836970574254,
664
+ "loss": 0.8262,
665
+ "step": 108
666
+ },
667
+ {
668
+ "epoch": 0.18,
669
+ "learning_rate": 0.00019668204261676618,
670
+ "loss": 0.7775,
671
+ "step": 109
672
+ },
673
+ {
674
+ "epoch": 0.18,
675
+ "learning_rate": 0.00019661505727555482,
676
+ "loss": 0.7647,
677
+ "step": 110
678
+ },
679
+ {
680
+ "epoch": 0.18,
681
+ "learning_rate": 0.00019654741413817296,
682
+ "loss": 0.8048,
683
+ "step": 111
684
+ },
685
+ {
686
+ "epoch": 0.18,
687
+ "learning_rate": 0.0001964791136651637,
688
+ "loss": 0.8262,
689
+ "step": 112
690
+ },
691
+ {
692
+ "epoch": 0.19,
693
+ "learning_rate": 0.00019641015632154552,
694
+ "loss": 0.8083,
695
+ "step": 113
696
+ },
697
+ {
698
+ "epoch": 0.19,
699
+ "learning_rate": 0.00019634054257680923,
700
+ "loss": 0.7768,
701
+ "step": 114
702
+ },
703
+ {
704
+ "epoch": 0.19,
705
+ "learning_rate": 0.00019627027290491458,
706
+ "loss": 0.794,
707
+ "step": 115
708
+ },
709
+ {
710
+ "epoch": 0.19,
711
+ "learning_rate": 0.0001961993477842873,
712
+ "loss": 0.8039,
713
+ "step": 116
714
+ },
715
+ {
716
+ "epoch": 0.19,
717
+ "learning_rate": 0.00019612776769781554,
718
+ "loss": 0.7718,
719
+ "step": 117
720
+ },
721
+ {
722
+ "epoch": 0.19,
723
+ "learning_rate": 0.0001960555331328468,
724
+ "loss": 0.7891,
725
+ "step": 118
726
+ },
727
+ {
728
+ "epoch": 0.2,
729
+ "learning_rate": 0.00019598264458118458,
730
+ "loss": 0.87,
731
+ "step": 119
732
+ },
733
+ {
734
+ "epoch": 0.2,
735
+ "learning_rate": 0.00019590910253908494,
736
+ "loss": 0.8009,
737
+ "step": 120
738
+ },
739
+ {
740
+ "epoch": 0.2,
741
+ "learning_rate": 0.00019583490750725325,
742
+ "loss": 0.7793,
743
+ "step": 121
744
+ },
745
+ {
746
+ "epoch": 0.2,
747
+ "learning_rate": 0.0001957600599908406,
748
+ "loss": 0.755,
749
+ "step": 122
750
+ },
751
+ {
752
+ "epoch": 0.2,
753
+ "learning_rate": 0.0001956845604994406,
754
+ "loss": 0.8324,
755
+ "step": 123
756
+ },
757
+ {
758
+ "epoch": 0.2,
759
+ "learning_rate": 0.00019560840954708565,
760
+ "loss": 0.801,
761
+ "step": 124
762
+ },
763
+ {
764
+ "epoch": 0.21,
765
+ "learning_rate": 0.00019553160765224372,
766
+ "loss": 0.7857,
767
+ "step": 125
768
+ },
769
+ {
770
+ "epoch": 0.21,
771
+ "learning_rate": 0.0001954541553378145,
772
+ "loss": 0.7975,
773
+ "step": 126
774
+ },
775
+ {
776
+ "epoch": 0.21,
777
+ "learning_rate": 0.0001953760531311262,
778
+ "loss": 0.7837,
779
+ "step": 127
780
+ },
781
+ {
782
+ "epoch": 0.21,
783
+ "learning_rate": 0.0001952973015639316,
784
+ "loss": 0.7926,
785
+ "step": 128
786
+ },
787
+ {
788
+ "epoch": 0.21,
789
+ "learning_rate": 0.0001952179011724047,
790
+ "loss": 0.7713,
791
+ "step": 129
792
+ },
793
+ {
794
+ "epoch": 0.21,
795
+ "learning_rate": 0.00019513785249713697,
796
+ "loss": 0.7883,
797
+ "step": 130
798
+ },
799
+ {
800
+ "epoch": 0.22,
801
+ "learning_rate": 0.00019505715608313359,
802
+ "loss": 0.8101,
803
+ "step": 131
804
+ },
805
+ {
806
+ "epoch": 0.22,
807
+ "learning_rate": 0.00019497581247980992,
808
+ "loss": 0.8487,
809
+ "step": 132
810
+ },
811
+ {
812
+ "epoch": 0.22,
813
+ "learning_rate": 0.0001948938222409876,
814
+ "loss": 0.7866,
815
+ "step": 133
816
+ },
817
+ {
818
+ "epoch": 0.22,
819
+ "learning_rate": 0.00019481118592489086,
820
+ "loss": 0.8004,
821
+ "step": 134
822
+ },
823
+ {
824
+ "epoch": 0.22,
825
+ "learning_rate": 0.00019472790409414266,
826
+ "loss": 0.8064,
827
+ "step": 135
828
+ },
829
+ {
830
+ "epoch": 0.22,
831
+ "learning_rate": 0.00019464397731576094,
832
+ "loss": 0.8209,
833
+ "step": 136
834
+ },
835
+ {
836
+ "epoch": 0.23,
837
+ "learning_rate": 0.00019455940616115472,
838
+ "loss": 0.8289,
839
+ "step": 137
840
+ },
841
+ {
842
+ "epoch": 0.23,
843
+ "learning_rate": 0.00019447419120612017,
844
+ "loss": 0.7755,
845
+ "step": 138
846
+ },
847
+ {
848
+ "epoch": 0.23,
849
+ "learning_rate": 0.00019438833303083678,
850
+ "loss": 0.7701,
851
+ "step": 139
852
+ },
853
+ {
854
+ "epoch": 0.23,
855
+ "learning_rate": 0.00019430183221986325,
856
+ "loss": 0.7563,
857
+ "step": 140
858
+ },
859
+ {
860
+ "epoch": 0.23,
861
+ "learning_rate": 0.0001942146893621337,
862
+ "loss": 0.7749,
863
+ "step": 141
864
+ },
865
+ {
866
+ "epoch": 0.23,
867
+ "learning_rate": 0.00019412690505095365,
868
+ "loss": 0.8033,
869
+ "step": 142
870
+ },
871
+ {
872
+ "epoch": 0.24,
873
+ "learning_rate": 0.0001940384798839957,
874
+ "loss": 0.8116,
875
+ "step": 143
876
+ },
877
+ {
878
+ "epoch": 0.24,
879
+ "learning_rate": 0.00019394941446329583,
880
+ "loss": 0.7957,
881
+ "step": 144
882
+ },
883
+ {
884
+ "epoch": 0.24,
885
+ "learning_rate": 0.0001938597093952491,
886
+ "loss": 0.7991,
887
+ "step": 145
888
+ },
889
+ {
890
+ "epoch": 0.24,
891
+ "learning_rate": 0.00019376936529060554,
892
+ "loss": 0.7871,
893
+ "step": 146
894
+ },
895
+ {
896
+ "epoch": 0.24,
897
+ "learning_rate": 0.000193678382764466,
898
+ "loss": 0.8345,
899
+ "step": 147
900
+ },
901
+ {
902
+ "epoch": 0.24,
903
+ "learning_rate": 0.00019358676243627808,
904
+ "loss": 0.7805,
905
+ "step": 148
906
+ },
907
+ {
908
+ "epoch": 0.25,
909
+ "learning_rate": 0.00019349450492983164,
910
+ "loss": 0.8089,
911
+ "step": 149
912
+ },
913
+ {
914
+ "epoch": 0.25,
915
+ "learning_rate": 0.0001934016108732548,
916
+ "loss": 0.7563,
917
+ "step": 150
918
+ },
919
+ {
920
+ "epoch": 0.25,
921
+ "learning_rate": 0.00019330808089900963,
922
+ "loss": 0.7729,
923
+ "step": 151
924
+ },
925
+ {
926
+ "epoch": 0.25,
927
+ "learning_rate": 0.00019321391564388775,
928
+ "loss": 0.8198,
929
+ "step": 152
930
+ },
931
+ {
932
+ "epoch": 0.25,
933
+ "eval_loss": 0.8079261183738708,
934
+ "eval_runtime": 85.1654,
935
+ "eval_samples_per_second": 4.039,
936
+ "eval_steps_per_second": 2.02,
937
+ "step": 152
938
+ },
939
+ {
940
+ "epoch": 0.25,
941
+ "learning_rate": 0.00019311911574900598,
942
+ "loss": 0.7712,
943
+ "step": 153
944
+ },
945
+ {
946
+ "epoch": 0.25,
947
+ "learning_rate": 0.00019302368185980217,
948
+ "loss": 0.7539,
949
+ "step": 154
950
+ },
951
+ {
952
+ "epoch": 0.26,
953
+ "learning_rate": 0.00019292761462603056,
954
+ "loss": 0.7872,
955
+ "step": 155
956
+ },
957
+ {
958
+ "epoch": 0.26,
959
+ "learning_rate": 0.00019283091470175754,
960
+ "loss": 0.7657,
961
+ "step": 156
962
+ },
963
+ {
964
+ "epoch": 0.26,
965
+ "learning_rate": 0.00019273358274535704,
966
+ "loss": 0.7503,
967
+ "step": 157
968
+ },
969
+ {
970
+ "epoch": 0.26,
971
+ "learning_rate": 0.00019263561941950622,
972
+ "loss": 0.7753,
973
+ "step": 158
974
+ },
975
+ {
976
+ "epoch": 0.26,
977
+ "learning_rate": 0.0001925370253911808,
978
+ "loss": 0.8214,
979
+ "step": 159
980
+ },
981
+ {
982
+ "epoch": 0.26,
983
+ "learning_rate": 0.00019243780133165067,
984
+ "loss": 0.839,
985
+ "step": 160
986
+ },
987
+ {
988
+ "epoch": 0.27,
989
+ "learning_rate": 0.00019233794791647516,
990
+ "loss": 0.7858,
991
+ "step": 161
992
+ },
993
+ {
994
+ "epoch": 0.27,
995
+ "learning_rate": 0.00019223746582549853,
996
+ "loss": 0.8081,
997
+ "step": 162
998
+ },
999
+ {
1000
+ "epoch": 0.27,
1001
+ "learning_rate": 0.0001921363557428454,
1002
+ "loss": 0.7762,
1003
+ "step": 163
1004
+ },
1005
+ {
1006
+ "epoch": 0.27,
1007
+ "learning_rate": 0.00019203461835691594,
1008
+ "loss": 0.7837,
1009
+ "step": 164
1010
+ },
1011
+ {
1012
+ "epoch": 0.27,
1013
+ "learning_rate": 0.00019193225436038133,
1014
+ "loss": 0.816,
1015
+ "step": 165
1016
+ },
1017
+ {
1018
+ "epoch": 0.27,
1019
+ "learning_rate": 0.00019182926445017893,
1020
+ "loss": 0.7962,
1021
+ "step": 166
1022
+ },
1023
+ {
1024
+ "epoch": 0.28,
1025
+ "learning_rate": 0.0001917256493275076,
1026
+ "loss": 0.8454,
1027
+ "step": 167
1028
+ },
1029
+ {
1030
+ "epoch": 0.28,
1031
+ "learning_rate": 0.00019162140969782292,
1032
+ "loss": 0.7748,
1033
+ "step": 168
1034
+ },
1035
+ {
1036
+ "epoch": 0.28,
1037
+ "learning_rate": 0.00019151654627083236,
1038
+ "loss": 0.782,
1039
+ "step": 169
1040
+ },
1041
+ {
1042
+ "epoch": 0.28,
1043
+ "learning_rate": 0.00019141105976049053,
1044
+ "loss": 0.8049,
1045
+ "step": 170
1046
+ },
1047
+ {
1048
+ "epoch": 0.28,
1049
+ "learning_rate": 0.0001913049508849942,
1050
+ "loss": 0.7263,
1051
+ "step": 171
1052
+ },
1053
+ {
1054
+ "epoch": 0.28,
1055
+ "learning_rate": 0.00019119822036677738,
1056
+ "loss": 0.8035,
1057
+ "step": 172
1058
+ },
1059
+ {
1060
+ "epoch": 0.29,
1061
+ "learning_rate": 0.00019109086893250664,
1062
+ "loss": 0.7649,
1063
+ "step": 173
1064
+ },
1065
+ {
1066
+ "epoch": 0.29,
1067
+ "learning_rate": 0.000190982897313076,
1068
+ "loss": 0.8311,
1069
+ "step": 174
1070
+ },
1071
+ {
1072
+ "epoch": 0.29,
1073
+ "learning_rate": 0.0001908743062436018,
1074
+ "loss": 0.8503,
1075
+ "step": 175
1076
+ },
1077
+ {
1078
+ "epoch": 0.29,
1079
+ "learning_rate": 0.00019076509646341806,
1080
+ "loss": 0.7664,
1081
+ "step": 176
1082
+ },
1083
+ {
1084
+ "epoch": 0.29,
1085
+ "learning_rate": 0.00019065526871607112,
1086
+ "loss": 0.805,
1087
+ "step": 177
1088
+ },
1089
+ {
1090
+ "epoch": 0.29,
1091
+ "learning_rate": 0.00019054482374931467,
1092
+ "loss": 0.7798,
1093
+ "step": 178
1094
+ },
1095
+ {
1096
+ "epoch": 0.29,
1097
+ "learning_rate": 0.00019043376231510484,
1098
+ "loss": 0.804,
1099
+ "step": 179
1100
+ },
1101
+ {
1102
+ "epoch": 0.3,
1103
+ "learning_rate": 0.0001903220851695948,
1104
+ "loss": 0.8074,
1105
+ "step": 180
1106
+ },
1107
+ {
1108
+ "epoch": 0.3,
1109
+ "learning_rate": 0.0001902097930731298,
1110
+ "loss": 0.8134,
1111
+ "step": 181
1112
+ },
1113
+ {
1114
+ "epoch": 0.3,
1115
+ "learning_rate": 0.0001900968867902419,
1116
+ "loss": 0.7793,
1117
+ "step": 182
1118
+ },
1119
+ {
1120
+ "epoch": 0.3,
1121
+ "learning_rate": 0.00018998336708964488,
1122
+ "loss": 0.793,
1123
+ "step": 183
1124
+ },
1125
+ {
1126
+ "epoch": 0.3,
1127
+ "learning_rate": 0.00018986923474422884,
1128
+ "loss": 0.781,
1129
+ "step": 184
1130
+ },
1131
+ {
1132
+ "epoch": 0.3,
1133
+ "learning_rate": 0.00018975449053105505,
1134
+ "loss": 0.858,
1135
+ "step": 185
1136
+ },
1137
+ {
1138
+ "epoch": 0.31,
1139
+ "learning_rate": 0.0001896391352313506,
1140
+ "loss": 0.7909,
1141
+ "step": 186
1142
+ },
1143
+ {
1144
+ "epoch": 0.31,
1145
+ "learning_rate": 0.00018952316963050328,
1146
+ "loss": 0.7954,
1147
+ "step": 187
1148
+ },
1149
+ {
1150
+ "epoch": 0.31,
1151
+ "learning_rate": 0.0001894065945180558,
1152
+ "loss": 0.7544,
1153
+ "step": 188
1154
+ },
1155
+ {
1156
+ "epoch": 0.31,
1157
+ "learning_rate": 0.00018928941068770093,
1158
+ "loss": 0.7575,
1159
+ "step": 189
1160
+ },
1161
+ {
1162
+ "epoch": 0.31,
1163
+ "learning_rate": 0.0001891716189372757,
1164
+ "loss": 0.7981,
1165
+ "step": 190
1166
+ },
1167
+ {
1168
+ "epoch": 0.31,
1169
+ "learning_rate": 0.00018905322006875617,
1170
+ "loss": 0.7997,
1171
+ "step": 191
1172
+ },
1173
+ {
1174
+ "epoch": 0.32,
1175
+ "learning_rate": 0.0001889342148882519,
1176
+ "loss": 0.7553,
1177
+ "step": 192
1178
+ },
1179
+ {
1180
+ "epoch": 0.32,
1181
+ "learning_rate": 0.0001888146042060005,
1182
+ "loss": 0.8128,
1183
+ "step": 193
1184
+ },
1185
+ {
1186
+ "epoch": 0.32,
1187
+ "learning_rate": 0.00018869438883636214,
1188
+ "loss": 0.8158,
1189
+ "step": 194
1190
+ },
1191
+ {
1192
+ "epoch": 0.32,
1193
+ "learning_rate": 0.00018857356959781378,
1194
+ "loss": 0.8029,
1195
+ "step": 195
1196
+ },
1197
+ {
1198
+ "epoch": 0.32,
1199
+ "learning_rate": 0.000188452147312944,
1200
+ "loss": 0.788,
1201
+ "step": 196
1202
+ },
1203
+ {
1204
+ "epoch": 0.32,
1205
+ "learning_rate": 0.00018833012280844699,
1206
+ "loss": 0.785,
1207
+ "step": 197
1208
+ },
1209
+ {
1210
+ "epoch": 0.33,
1211
+ "learning_rate": 0.0001882074969151172,
1212
+ "loss": 0.7955,
1213
+ "step": 198
1214
+ },
1215
+ {
1216
+ "epoch": 0.33,
1217
+ "learning_rate": 0.00018808427046784366,
1218
+ "loss": 0.8033,
1219
+ "step": 199
1220
+ },
1221
+ {
1222
+ "epoch": 0.33,
1223
+ "learning_rate": 0.000187960444305604,
1224
+ "loss": 0.7744,
1225
+ "step": 200
1226
+ },
1227
+ {
1228
+ "epoch": 0.33,
1229
+ "learning_rate": 0.0001878360192714592,
1230
+ "loss": 0.8084,
1231
+ "step": 201
1232
+ },
1233
+ {
1234
+ "epoch": 0.33,
1235
+ "learning_rate": 0.00018771099621254746,
1236
+ "loss": 0.8271,
1237
+ "step": 202
1238
+ },
1239
+ {
1240
+ "epoch": 0.33,
1241
+ "learning_rate": 0.00018758537598007868,
1242
+ "loss": 0.8153,
1243
+ "step": 203
1244
+ },
1245
+ {
1246
+ "epoch": 0.34,
1247
+ "learning_rate": 0.0001874591594293285,
1248
+ "loss": 0.7638,
1249
+ "step": 204
1250
+ },
1251
+ {
1252
+ "epoch": 0.34,
1253
+ "learning_rate": 0.0001873323474196326,
1254
+ "loss": 0.7825,
1255
+ "step": 205
1256
+ },
1257
+ {
1258
+ "epoch": 0.34,
1259
+ "learning_rate": 0.00018720494081438078,
1260
+ "loss": 0.7939,
1261
+ "step": 206
1262
+ },
1263
+ {
1264
+ "epoch": 0.34,
1265
+ "learning_rate": 0.00018707694048101104,
1266
+ "loss": 0.8289,
1267
+ "step": 207
1268
+ },
1269
+ {
1270
+ "epoch": 0.34,
1271
+ "learning_rate": 0.00018694834729100386,
1272
+ "loss": 0.7978,
1273
+ "step": 208
1274
+ },
1275
+ {
1276
+ "epoch": 0.34,
1277
+ "learning_rate": 0.00018681916211987597,
1278
+ "loss": 0.7921,
1279
+ "step": 209
1280
+ },
1281
+ {
1282
+ "epoch": 0.35,
1283
+ "learning_rate": 0.00018668938584717471,
1284
+ "loss": 0.7849,
1285
+ "step": 210
1286
+ },
1287
+ {
1288
+ "epoch": 0.35,
1289
+ "learning_rate": 0.00018655901935647187,
1290
+ "loss": 0.7512,
1291
+ "step": 211
1292
+ },
1293
+ {
1294
+ "epoch": 0.35,
1295
+ "learning_rate": 0.00018642806353535754,
1296
+ "loss": 0.7452,
1297
+ "step": 212
1298
+ },
1299
+ {
1300
+ "epoch": 0.35,
1301
+ "learning_rate": 0.00018629651927543447,
1302
+ "loss": 0.7714,
1303
+ "step": 213
1304
+ },
1305
+ {
1306
+ "epoch": 0.35,
1307
+ "learning_rate": 0.00018616438747231148,
1308
+ "loss": 0.7974,
1309
+ "step": 214
1310
+ },
1311
+ {
1312
+ "epoch": 0.35,
1313
+ "learning_rate": 0.00018603166902559783,
1314
+ "loss": 0.7938,
1315
+ "step": 215
1316
+ },
1317
+ {
1318
+ "epoch": 0.36,
1319
+ "learning_rate": 0.00018589836483889687,
1320
+ "loss": 0.7861,
1321
+ "step": 216
1322
+ },
1323
+ {
1324
+ "epoch": 0.36,
1325
+ "learning_rate": 0.00018576447581979984,
1326
+ "loss": 0.7472,
1327
+ "step": 217
1328
+ },
1329
+ {
1330
+ "epoch": 0.36,
1331
+ "learning_rate": 0.0001856300028798798,
1332
+ "loss": 0.7789,
1333
+ "step": 218
1334
+ },
1335
+ {
1336
+ "epoch": 0.36,
1337
+ "learning_rate": 0.0001854949469346854,
1338
+ "loss": 0.7047,
1339
+ "step": 219
1340
+ },
1341
+ {
1342
+ "epoch": 0.36,
1343
+ "learning_rate": 0.00018535930890373466,
1344
+ "loss": 0.7304,
1345
+ "step": 220
1346
+ },
1347
+ {
1348
+ "epoch": 0.36,
1349
+ "learning_rate": 0.00018522308971050865,
1350
+ "loss": 0.7701,
1351
+ "step": 221
1352
+ },
1353
+ {
1354
+ "epoch": 0.37,
1355
+ "learning_rate": 0.00018508629028244519,
1356
+ "loss": 0.8023,
1357
+ "step": 222
1358
+ },
1359
+ {
1360
+ "epoch": 0.37,
1361
+ "learning_rate": 0.00018494891155093274,
1362
+ "loss": 0.8144,
1363
+ "step": 223
1364
+ },
1365
+ {
1366
+ "epoch": 0.37,
1367
+ "learning_rate": 0.0001848109544513037,
1368
+ "loss": 0.7951,
1369
+ "step": 224
1370
+ },
1371
+ {
1372
+ "epoch": 0.37,
1373
+ "learning_rate": 0.00018467241992282843,
1374
+ "loss": 0.7478,
1375
+ "step": 225
1376
+ },
1377
+ {
1378
+ "epoch": 0.37,
1379
+ "learning_rate": 0.00018453330890870855,
1380
+ "loss": 0.8309,
1381
+ "step": 226
1382
+ },
1383
+ {
1384
+ "epoch": 0.37,
1385
+ "learning_rate": 0.0001843936223560707,
1386
+ "loss": 0.7643,
1387
+ "step": 227
1388
+ },
1389
+ {
1390
+ "epoch": 0.38,
1391
+ "learning_rate": 0.00018425336121596,
1392
+ "loss": 0.7689,
1393
+ "step": 228
1394
+ },
1395
+ {
1396
+ "epoch": 0.38,
1397
+ "learning_rate": 0.00018411252644333362,
1398
+ "loss": 0.7866,
1399
+ "step": 229
1400
+ },
1401
+ {
1402
+ "epoch": 0.38,
1403
+ "learning_rate": 0.00018397111899705419,
1404
+ "loss": 0.7165,
1405
+ "step": 230
1406
+ },
1407
+ {
1408
+ "epoch": 0.38,
1409
+ "learning_rate": 0.00018382913983988348,
1410
+ "loss": 0.8189,
1411
+ "step": 231
1412
+ },
1413
+ {
1414
+ "epoch": 0.38,
1415
+ "learning_rate": 0.00018368658993847566,
1416
+ "loss": 0.7934,
1417
+ "step": 232
1418
+ },
1419
+ {
1420
+ "epoch": 0.38,
1421
+ "learning_rate": 0.00018354347026337066,
1422
+ "loss": 0.7668,
1423
+ "step": 233
1424
+ },
1425
+ {
1426
+ "epoch": 0.39,
1427
+ "learning_rate": 0.0001833997817889878,
1428
+ "loss": 0.7278,
1429
+ "step": 234
1430
+ },
1431
+ {
1432
+ "epoch": 0.39,
1433
+ "learning_rate": 0.000183255525493619,
1434
+ "loss": 0.7661,
1435
+ "step": 235
1436
+ },
1437
+ {
1438
+ "epoch": 0.39,
1439
+ "learning_rate": 0.0001831107023594221,
1440
+ "loss": 0.8165,
1441
+ "step": 236
1442
+ },
1443
+ {
1444
+ "epoch": 0.39,
1445
+ "learning_rate": 0.00018296531337241425,
1446
+ "loss": 0.8411,
1447
+ "step": 237
1448
+ },
1449
+ {
1450
+ "epoch": 0.39,
1451
+ "learning_rate": 0.00018281935952246518,
1452
+ "loss": 0.7806,
1453
+ "step": 238
1454
+ },
1455
+ {
1456
+ "epoch": 0.39,
1457
+ "learning_rate": 0.0001826728418032904,
1458
+ "loss": 0.7821,
1459
+ "step": 239
1460
+ },
1461
+ {
1462
+ "epoch": 0.4,
1463
+ "learning_rate": 0.00018252576121244456,
1464
+ "loss": 0.7878,
1465
+ "step": 240
1466
+ },
1467
+ {
1468
+ "epoch": 0.4,
1469
+ "learning_rate": 0.00018237811875131444,
1470
+ "loss": 0.7586,
1471
+ "step": 241
1472
+ },
1473
+ {
1474
+ "epoch": 0.4,
1475
+ "learning_rate": 0.00018222991542511246,
1476
+ "loss": 0.7855,
1477
+ "step": 242
1478
+ },
1479
+ {
1480
+ "epoch": 0.4,
1481
+ "learning_rate": 0.00018208115224286947,
1482
+ "loss": 0.7881,
1483
+ "step": 243
1484
+ },
1485
+ {
1486
+ "epoch": 0.4,
1487
+ "learning_rate": 0.0001819318302174281,
1488
+ "loss": 0.7394,
1489
+ "step": 244
1490
+ },
1491
+ {
1492
+ "epoch": 0.4,
1493
+ "learning_rate": 0.00018178195036543592,
1494
+ "loss": 0.7526,
1495
+ "step": 245
1496
+ },
1497
+ {
1498
+ "epoch": 0.41,
1499
+ "learning_rate": 0.00018163151370733838,
1500
+ "loss": 0.7969,
1501
+ "step": 246
1502
+ },
1503
+ {
1504
+ "epoch": 0.41,
1505
+ "learning_rate": 0.00018148052126737177,
1506
+ "loss": 0.7877,
1507
+ "step": 247
1508
+ },
1509
+ {
1510
+ "epoch": 0.41,
1511
+ "learning_rate": 0.00018132897407355657,
1512
+ "loss": 0.7849,
1513
+ "step": 248
1514
+ },
1515
+ {
1516
+ "epoch": 0.41,
1517
+ "learning_rate": 0.00018117687315769007,
1518
+ "loss": 0.7871,
1519
+ "step": 249
1520
+ },
1521
+ {
1522
+ "epoch": 0.41,
1523
+ "learning_rate": 0.00018102421955533974,
1524
+ "loss": 0.7423,
1525
+ "step": 250
1526
+ },
1527
+ {
1528
+ "epoch": 0.41,
1529
+ "learning_rate": 0.00018087101430583577,
1530
+ "loss": 0.7658,
1531
+ "step": 251
1532
+ },
1533
+ {
1534
+ "epoch": 0.42,
1535
+ "learning_rate": 0.00018071725845226436,
1536
+ "loss": 0.7778,
1537
+ "step": 252
1538
+ },
1539
+ {
1540
+ "epoch": 0.42,
1541
+ "learning_rate": 0.0001805629530414604,
1542
+ "loss": 0.7429,
1543
+ "step": 253
1544
+ },
1545
+ {
1546
+ "epoch": 0.42,
1547
+ "learning_rate": 0.0001804080991240003,
1548
+ "loss": 0.7622,
1549
+ "step": 254
1550
+ },
1551
+ {
1552
+ "epoch": 0.42,
1553
+ "learning_rate": 0.00018025269775419507,
1554
+ "loss": 0.8222,
1555
+ "step": 255
1556
+ },
1557
+ {
1558
+ "epoch": 0.42,
1559
+ "learning_rate": 0.000180096749990083,
1560
+ "loss": 0.7611,
1561
+ "step": 256
1562
+ },
1563
+ {
1564
+ "epoch": 0.42,
1565
+ "learning_rate": 0.00017994025689342235,
1566
+ "loss": 0.7496,
1567
+ "step": 257
1568
+ },
1569
+ {
1570
+ "epoch": 0.43,
1571
+ "learning_rate": 0.00017978321952968434,
1572
+ "loss": 0.7821,
1573
+ "step": 258
1574
+ },
1575
+ {
1576
+ "epoch": 0.43,
1577
+ "learning_rate": 0.00017962563896804578,
1578
+ "loss": 0.732,
1579
+ "step": 259
1580
+ },
1581
+ {
1582
+ "epoch": 0.43,
1583
+ "learning_rate": 0.00017946751628138174,
1584
+ "loss": 0.7608,
1585
+ "step": 260
1586
+ },
1587
+ {
1588
+ "epoch": 0.43,
1589
+ "learning_rate": 0.00017930885254625832,
1590
+ "loss": 0.7711,
1591
+ "step": 261
1592
+ },
1593
+ {
1594
+ "epoch": 0.43,
1595
+ "learning_rate": 0.00017914964884292544,
1596
+ "loss": 0.769,
1597
+ "step": 262
1598
+ },
1599
+ {
1600
+ "epoch": 0.43,
1601
+ "learning_rate": 0.0001789899062553091,
1602
+ "loss": 0.7177,
1603
+ "step": 263
1604
+ },
1605
+ {
1606
+ "epoch": 0.43,
1607
+ "learning_rate": 0.0001788296258710045,
1608
+ "loss": 0.7931,
1609
+ "step": 264
1610
+ },
1611
+ {
1612
+ "epoch": 0.44,
1613
+ "learning_rate": 0.00017866880878126824,
1614
+ "loss": 0.745,
1615
+ "step": 265
1616
+ },
1617
+ {
1618
+ "epoch": 0.44,
1619
+ "learning_rate": 0.0001785074560810111,
1620
+ "loss": 0.8056,
1621
+ "step": 266
1622
+ },
1623
+ {
1624
+ "epoch": 0.44,
1625
+ "learning_rate": 0.00017834556886879045,
1626
+ "loss": 0.7687,
1627
+ "step": 267
1628
+ },
1629
+ {
1630
+ "epoch": 0.44,
1631
+ "learning_rate": 0.000178183148246803,
1632
+ "loss": 0.7665,
1633
+ "step": 268
1634
+ },
1635
+ {
1636
+ "epoch": 0.44,
1637
+ "learning_rate": 0.00017802019532087694,
1638
+ "loss": 0.7694,
1639
+ "step": 269
1640
+ },
1641
+ {
1642
+ "epoch": 0.44,
1643
+ "learning_rate": 0.00017785671120046473,
1644
+ "loss": 0.7823,
1645
+ "step": 270
1646
+ },
1647
+ {
1648
+ "epoch": 0.45,
1649
+ "learning_rate": 0.00017769269699863542,
1650
+ "loss": 0.7712,
1651
+ "step": 271
1652
+ },
1653
+ {
1654
+ "epoch": 0.45,
1655
+ "learning_rate": 0.00017752815383206705,
1656
+ "loss": 0.7356,
1657
+ "step": 272
1658
+ },
1659
+ {
1660
+ "epoch": 0.45,
1661
+ "learning_rate": 0.00017736308282103908,
1662
+ "loss": 0.7644,
1663
+ "step": 273
1664
+ },
1665
+ {
1666
+ "epoch": 0.45,
1667
+ "learning_rate": 0.0001771974850894248,
1668
+ "loss": 0.7623,
1669
+ "step": 274
1670
+ },
1671
+ {
1672
+ "epoch": 0.45,
1673
+ "learning_rate": 0.00017703136176468355,
1674
+ "loss": 0.7771,
1675
+ "step": 275
1676
+ },
1677
+ {
1678
+ "epoch": 0.45,
1679
+ "learning_rate": 0.0001768647139778532,
1680
+ "loss": 0.832,
1681
+ "step": 276
1682
+ },
1683
+ {
1684
+ "epoch": 0.46,
1685
+ "learning_rate": 0.00017669754286354241,
1686
+ "loss": 0.7733,
1687
+ "step": 277
1688
+ },
1689
+ {
1690
+ "epoch": 0.46,
1691
+ "learning_rate": 0.00017652984955992277,
1692
+ "loss": 0.7669,
1693
+ "step": 278
1694
+ },
1695
+ {
1696
+ "epoch": 0.46,
1697
+ "learning_rate": 0.00017636163520872122,
1698
+ "loss": 0.8019,
1699
+ "step": 279
1700
+ },
1701
+ {
1702
+ "epoch": 0.46,
1703
+ "learning_rate": 0.0001761929009552122,
1704
+ "loss": 0.7812,
1705
+ "step": 280
1706
+ },
1707
+ {
1708
+ "epoch": 0.46,
1709
+ "learning_rate": 0.0001760236479482099,
1710
+ "loss": 0.7761,
1711
+ "step": 281
1712
+ },
1713
+ {
1714
+ "epoch": 0.46,
1715
+ "learning_rate": 0.00017585387734006034,
1716
+ "loss": 0.835,
1717
+ "step": 282
1718
+ },
1719
+ {
1720
+ "epoch": 0.47,
1721
+ "learning_rate": 0.00017568359028663364,
1722
+ "loss": 0.7895,
1723
+ "step": 283
1724
+ },
1725
+ {
1726
+ "epoch": 0.47,
1727
+ "learning_rate": 0.00017551278794731607,
1728
+ "loss": 0.7483,
1729
+ "step": 284
1730
+ },
1731
+ {
1732
+ "epoch": 0.47,
1733
+ "learning_rate": 0.0001753414714850022,
1734
+ "loss": 0.7672,
1735
+ "step": 285
1736
+ },
1737
+ {
1738
+ "epoch": 0.47,
1739
+ "learning_rate": 0.00017516964206608696,
1740
+ "loss": 0.777,
1741
+ "step": 286
1742
+ },
1743
+ {
1744
+ "epoch": 0.47,
1745
+ "learning_rate": 0.00017499730086045767,
1746
+ "loss": 0.8055,
1747
+ "step": 287
1748
+ },
1749
+ {
1750
+ "epoch": 0.47,
1751
+ "learning_rate": 0.00017482444904148617,
1752
+ "loss": 0.7941,
1753
+ "step": 288
1754
+ },
1755
+ {
1756
+ "epoch": 0.48,
1757
+ "learning_rate": 0.0001746510877860208,
1758
+ "loss": 0.7886,
1759
+ "step": 289
1760
+ },
1761
+ {
1762
+ "epoch": 0.48,
1763
+ "learning_rate": 0.0001744772182743782,
1764
+ "loss": 0.7651,
1765
+ "step": 290
1766
+ },
1767
+ {
1768
+ "epoch": 0.48,
1769
+ "learning_rate": 0.0001743028416903356,
1770
+ "loss": 0.7244,
1771
+ "step": 291
1772
+ },
1773
+ {
1774
+ "epoch": 0.48,
1775
+ "learning_rate": 0.00017412795922112253,
1776
+ "loss": 0.7792,
1777
+ "step": 292
1778
+ },
1779
+ {
1780
+ "epoch": 0.48,
1781
+ "learning_rate": 0.0001739525720574128,
1782
+ "loss": 0.7187,
1783
+ "step": 293
1784
+ },
1785
+ {
1786
+ "epoch": 0.48,
1787
+ "learning_rate": 0.0001737766813933164,
1788
+ "loss": 0.7578,
1789
+ "step": 294
1790
+ },
1791
+ {
1792
+ "epoch": 0.49,
1793
+ "learning_rate": 0.0001736002884263713,
1794
+ "loss": 0.7841,
1795
+ "step": 295
1796
+ },
1797
+ {
1798
+ "epoch": 0.49,
1799
+ "learning_rate": 0.00017342339435753553,
1800
+ "loss": 0.7706,
1801
+ "step": 296
1802
+ },
1803
+ {
1804
+ "epoch": 0.49,
1805
+ "learning_rate": 0.00017324600039117863,
1806
+ "loss": 0.7369,
1807
+ "step": 297
1808
+ },
1809
+ {
1810
+ "epoch": 0.49,
1811
+ "learning_rate": 0.00017306810773507376,
1812
+ "loss": 0.7624,
1813
+ "step": 298
1814
+ },
1815
+ {
1816
+ "epoch": 0.49,
1817
+ "learning_rate": 0.00017288971760038942,
1818
+ "loss": 0.7672,
1819
+ "step": 299
1820
+ },
1821
+ {
1822
+ "epoch": 0.49,
1823
+ "learning_rate": 0.00017271083120168102,
1824
+ "loss": 0.7777,
1825
+ "step": 300
1826
+ },
1827
+ {
1828
+ "epoch": 0.5,
1829
+ "learning_rate": 0.00017253144975688285,
1830
+ "loss": 0.7862,
1831
+ "step": 301
1832
+ },
1833
+ {
1834
+ "epoch": 0.5,
1835
+ "learning_rate": 0.00017235157448729967,
1836
+ "loss": 0.7499,
1837
+ "step": 302
1838
+ },
1839
+ {
1840
+ "epoch": 0.5,
1841
+ "learning_rate": 0.00017217120661759832,
1842
+ "loss": 0.7255,
1843
+ "step": 303
1844
+ },
1845
+ {
1846
+ "epoch": 0.5,
1847
+ "learning_rate": 0.0001719903473757996,
1848
+ "loss": 0.7993,
1849
+ "step": 304
1850
+ },
1851
+ {
1852
+ "epoch": 0.5,
1853
+ "eval_loss": 0.7903943061828613,
1854
+ "eval_runtime": 85.3888,
1855
+ "eval_samples_per_second": 4.029,
1856
+ "eval_steps_per_second": 2.014,
1857
+ "step": 304
1858
+ },
1859
+ {
1860
+ "epoch": 0.5,
1861
+ "learning_rate": 0.0001718089979932697,
1862
+ "loss": 0.7813,
1863
+ "step": 305
1864
+ },
1865
+ {
1866
+ "epoch": 0.5,
1867
+ "learning_rate": 0.0001716271597047119,
1868
+ "loss": 0.7613,
1869
+ "step": 306
1870
+ },
1871
+ {
1872
+ "epoch": 0.51,
1873
+ "learning_rate": 0.0001714448337481582,
1874
+ "loss": 0.7356,
1875
+ "step": 307
1876
+ },
1877
+ {
1878
+ "epoch": 0.51,
1879
+ "learning_rate": 0.0001712620213649608,
1880
+ "loss": 0.7831,
1881
+ "step": 308
1882
+ },
1883
+ {
1884
+ "epoch": 0.51,
1885
+ "learning_rate": 0.00017107872379978374,
1886
+ "loss": 0.7554,
1887
+ "step": 309
1888
+ },
1889
+ {
1890
+ "epoch": 0.51,
1891
+ "learning_rate": 0.00017089494230059432,
1892
+ "loss": 0.7791,
1893
+ "step": 310
1894
+ },
1895
+ {
1896
+ "epoch": 0.51,
1897
+ "learning_rate": 0.00017071067811865476,
1898
+ "loss": 0.7561,
1899
+ "step": 311
1900
+ },
1901
+ {
1902
+ "epoch": 0.51,
1903
+ "learning_rate": 0.0001705259325085135,
1904
+ "loss": 0.7705,
1905
+ "step": 312
1906
+ },
1907
+ {
1908
+ "epoch": 0.52,
1909
+ "learning_rate": 0.00017034070672799684,
1910
+ "loss": 0.7968,
1911
+ "step": 313
1912
+ },
1913
+ {
1914
+ "epoch": 0.52,
1915
+ "learning_rate": 0.00017015500203820022,
1916
+ "loss": 0.7709,
1917
+ "step": 314
1918
+ },
1919
+ {
1920
+ "epoch": 0.52,
1921
+ "learning_rate": 0.00016996881970347962,
1922
+ "loss": 0.7818,
1923
+ "step": 315
1924
+ },
1925
+ {
1926
+ "epoch": 0.52,
1927
+ "learning_rate": 0.0001697821609914432,
1928
+ "loss": 0.8448,
1929
+ "step": 316
1930
+ },
1931
+ {
1932
+ "epoch": 0.52,
1933
+ "learning_rate": 0.00016959502717294242,
1934
+ "loss": 0.7488,
1935
+ "step": 317
1936
+ },
1937
+ {
1938
+ "epoch": 0.52,
1939
+ "learning_rate": 0.0001694074195220634,
1940
+ "loss": 0.795,
1941
+ "step": 318
1942
+ },
1943
+ {
1944
+ "epoch": 0.53,
1945
+ "learning_rate": 0.0001692193393161184,
1946
+ "loss": 0.8145,
1947
+ "step": 319
1948
+ },
1949
+ {
1950
+ "epoch": 0.53,
1951
+ "learning_rate": 0.000169030787835637,
1952
+ "loss": 0.7788,
1953
+ "step": 320
1954
+ },
1955
+ {
1956
+ "epoch": 0.53,
1957
+ "learning_rate": 0.00016884176636435748,
1958
+ "loss": 0.7627,
1959
+ "step": 321
1960
+ },
1961
+ {
1962
+ "epoch": 0.53,
1963
+ "learning_rate": 0.00016865227618921788,
1964
+ "loss": 0.7744,
1965
+ "step": 322
1966
+ },
1967
+ {
1968
+ "epoch": 0.53,
1969
+ "learning_rate": 0.00016846231860034747,
1970
+ "loss": 0.7998,
1971
+ "step": 323
1972
+ },
1973
+ {
1974
+ "epoch": 0.53,
1975
+ "learning_rate": 0.00016827189489105788,
1976
+ "loss": 0.779,
1977
+ "step": 324
1978
+ },
1979
+ {
1980
+ "epoch": 0.54,
1981
+ "learning_rate": 0.00016808100635783423,
1982
+ "loss": 0.7434,
1983
+ "step": 325
1984
+ },
1985
+ {
1986
+ "epoch": 0.54,
1987
+ "learning_rate": 0.00016788965430032638,
1988
+ "loss": 0.717,
1989
+ "step": 326
1990
+ },
1991
+ {
1992
+ "epoch": 0.54,
1993
+ "learning_rate": 0.00016769784002134008,
1994
+ "loss": 0.729,
1995
+ "step": 327
1996
+ },
1997
+ {
1998
+ "epoch": 0.54,
1999
+ "learning_rate": 0.00016750556482682805,
2000
+ "loss": 0.7918,
2001
+ "step": 328
2002
+ },
2003
+ {
2004
+ "epoch": 0.54,
2005
+ "learning_rate": 0.0001673128300258812,
2006
+ "loss": 0.77,
2007
+ "step": 329
2008
+ },
2009
+ {
2010
+ "epoch": 0.54,
2011
+ "learning_rate": 0.00016711963693071943,
2012
+ "loss": 0.8369,
2013
+ "step": 330
2014
+ },
2015
+ {
2016
+ "epoch": 0.55,
2017
+ "learning_rate": 0.00016692598685668318,
2018
+ "loss": 0.8058,
2019
+ "step": 331
2020
+ },
2021
+ {
2022
+ "epoch": 0.55,
2023
+ "learning_rate": 0.00016673188112222394,
2024
+ "loss": 0.7752,
2025
+ "step": 332
2026
+ },
2027
+ {
2028
+ "epoch": 0.55,
2029
+ "learning_rate": 0.00016653732104889572,
2030
+ "loss": 0.728,
2031
+ "step": 333
2032
+ },
2033
+ {
2034
+ "epoch": 0.55,
2035
+ "learning_rate": 0.00016634230796134576,
2036
+ "loss": 0.7681,
2037
+ "step": 334
2038
+ },
2039
+ {
2040
+ "epoch": 0.55,
2041
+ "learning_rate": 0.0001661468431873056,
2042
+ "loss": 0.7822,
2043
+ "step": 335
2044
+ },
2045
+ {
2046
+ "epoch": 0.55,
2047
+ "learning_rate": 0.0001659509280575821,
2048
+ "loss": 0.766,
2049
+ "step": 336
2050
+ },
2051
+ {
2052
+ "epoch": 0.56,
2053
+ "learning_rate": 0.0001657545639060484,
2054
+ "loss": 0.7927,
2055
+ "step": 337
2056
+ },
2057
+ {
2058
+ "epoch": 0.56,
2059
+ "learning_rate": 0.0001655577520696346,
2060
+ "loss": 0.7923,
2061
+ "step": 338
2062
+ },
2063
+ {
2064
+ "epoch": 0.56,
2065
+ "learning_rate": 0.00016536049388831894,
2066
+ "loss": 0.7688,
2067
+ "step": 339
2068
+ },
2069
+ {
2070
+ "epoch": 0.56,
2071
+ "learning_rate": 0.00016516279070511854,
2072
+ "loss": 0.7411,
2073
+ "step": 340
2074
+ },
2075
+ {
2076
+ "epoch": 0.56,
2077
+ "learning_rate": 0.0001649646438660803,
2078
+ "loss": 0.777,
2079
+ "step": 341
2080
+ },
2081
+ {
2082
+ "epoch": 0.56,
2083
+ "learning_rate": 0.00016476605472027172,
2084
+ "loss": 0.8018,
2085
+ "step": 342
2086
+ },
2087
+ {
2088
+ "epoch": 0.57,
2089
+ "learning_rate": 0.0001645670246197716,
2090
+ "loss": 0.7639,
2091
+ "step": 343
2092
+ },
2093
+ {
2094
+ "epoch": 0.57,
2095
+ "learning_rate": 0.00016436755491966115,
2096
+ "loss": 0.7529,
2097
+ "step": 344
2098
+ },
2099
+ {
2100
+ "epoch": 0.57,
2101
+ "learning_rate": 0.00016416764697801438,
2102
+ "loss": 0.7715,
2103
+ "step": 345
2104
+ },
2105
+ {
2106
+ "epoch": 0.57,
2107
+ "learning_rate": 0.00016396730215588915,
2108
+ "loss": 0.7675,
2109
+ "step": 346
2110
+ },
2111
+ {
2112
+ "epoch": 0.57,
2113
+ "learning_rate": 0.00016376652181731769,
2114
+ "loss": 0.7334,
2115
+ "step": 347
2116
+ },
2117
+ {
2118
+ "epoch": 0.57,
2119
+ "learning_rate": 0.0001635653073292975,
2120
+ "loss": 0.8244,
2121
+ "step": 348
2122
+ },
2123
+ {
2124
+ "epoch": 0.57,
2125
+ "learning_rate": 0.00016336366006178187,
2126
+ "loss": 0.7749,
2127
+ "step": 349
2128
+ },
2129
+ {
2130
+ "epoch": 0.58,
2131
+ "learning_rate": 0.0001631615813876707,
2132
+ "loss": 0.7767,
2133
+ "step": 350
2134
+ },
2135
+ {
2136
+ "epoch": 0.58,
2137
+ "learning_rate": 0.00016295907268280109,
2138
+ "loss": 0.7952,
2139
+ "step": 351
2140
+ },
2141
+ {
2142
+ "epoch": 0.58,
2143
+ "learning_rate": 0.0001627561353259379,
2144
+ "loss": 0.7624,
2145
+ "step": 352
2146
+ },
2147
+ {
2148
+ "epoch": 0.58,
2149
+ "learning_rate": 0.00016255277069876454,
2150
+ "loss": 0.7682,
2151
+ "step": 353
2152
+ },
2153
+ {
2154
+ "epoch": 0.58,
2155
+ "learning_rate": 0.00016234898018587337,
2156
+ "loss": 0.7176,
2157
+ "step": 354
2158
+ },
2159
+ {
2160
+ "epoch": 0.58,
2161
+ "learning_rate": 0.00016214476517475637,
2162
+ "loss": 0.7674,
2163
+ "step": 355
2164
+ },
2165
+ {
2166
+ "epoch": 0.59,
2167
+ "learning_rate": 0.00016194012705579572,
2168
+ "loss": 0.8107,
2169
+ "step": 356
2170
+ },
2171
+ {
2172
+ "epoch": 0.59,
2173
+ "learning_rate": 0.00016173506722225428,
2174
+ "loss": 0.7778,
2175
+ "step": 357
2176
+ },
2177
+ {
2178
+ "epoch": 0.59,
2179
+ "learning_rate": 0.00016152958707026614,
2180
+ "loss": 0.7688,
2181
+ "step": 358
2182
+ },
2183
+ {
2184
+ "epoch": 0.59,
2185
+ "learning_rate": 0.00016132368799882704,
2186
+ "loss": 0.7777,
2187
+ "step": 359
2188
+ },
2189
+ {
2190
+ "epoch": 0.59,
2191
+ "learning_rate": 0.00016111737140978494,
2192
+ "loss": 0.7688,
2193
+ "step": 360
2194
+ },
2195
+ {
2196
+ "epoch": 0.59,
2197
+ "learning_rate": 0.00016091063870783047,
2198
+ "loss": 0.737,
2199
+ "step": 361
2200
+ },
2201
+ {
2202
+ "epoch": 0.6,
2203
+ "learning_rate": 0.00016070349130048724,
2204
+ "loss": 0.7176,
2205
+ "step": 362
2206
+ },
2207
+ {
2208
+ "epoch": 0.6,
2209
+ "learning_rate": 0.00016049593059810248,
2210
+ "loss": 0.773,
2211
+ "step": 363
2212
+ },
2213
+ {
2214
+ "epoch": 0.6,
2215
+ "learning_rate": 0.00016028795801383718,
2216
+ "loss": 0.7305,
2217
+ "step": 364
2218
+ },
2219
+ {
2220
+ "epoch": 0.6,
2221
+ "learning_rate": 0.00016007957496365667,
2222
+ "loss": 0.7872,
2223
+ "step": 365
2224
+ },
2225
+ {
2226
+ "epoch": 0.6,
2227
+ "learning_rate": 0.0001598707828663209,
2228
+ "loss": 0.764,
2229
+ "step": 366
2230
+ },
2231
+ {
2232
+ "epoch": 0.6,
2233
+ "learning_rate": 0.00015966158314337472,
2234
+ "loss": 0.765,
2235
+ "step": 367
2236
+ },
2237
+ {
2238
+ "epoch": 0.61,
2239
+ "learning_rate": 0.00015945197721913833,
2240
+ "loss": 0.7292,
2241
+ "step": 368
2242
+ },
2243
+ {
2244
+ "epoch": 0.61,
2245
+ "learning_rate": 0.00015924196652069758,
2246
+ "loss": 0.7311,
2247
+ "step": 369
2248
+ },
2249
+ {
2250
+ "epoch": 0.61,
2251
+ "learning_rate": 0.00015903155247789404,
2252
+ "loss": 0.8094,
2253
+ "step": 370
2254
+ },
2255
+ {
2256
+ "epoch": 0.61,
2257
+ "learning_rate": 0.00015882073652331556,
2258
+ "loss": 0.7841,
2259
+ "step": 371
2260
+ },
2261
+ {
2262
+ "epoch": 0.61,
2263
+ "learning_rate": 0.00015860952009228625,
2264
+ "loss": 0.7609,
2265
+ "step": 372
2266
+ },
2267
+ {
2268
+ "epoch": 0.61,
2269
+ "learning_rate": 0.00015839790462285696,
2270
+ "loss": 0.7436,
2271
+ "step": 373
2272
+ },
2273
+ {
2274
+ "epoch": 0.62,
2275
+ "learning_rate": 0.0001581858915557953,
2276
+ "loss": 0.8029,
2277
+ "step": 374
2278
+ },
2279
+ {
2280
+ "epoch": 0.62,
2281
+ "learning_rate": 0.00015797348233457584,
2282
+ "loss": 0.7918,
2283
+ "step": 375
2284
+ },
2285
+ {
2286
+ "epoch": 0.62,
2287
+ "learning_rate": 0.0001577606784053705,
2288
+ "loss": 0.759,
2289
+ "step": 376
2290
+ },
2291
+ {
2292
+ "epoch": 0.62,
2293
+ "learning_rate": 0.0001575474812170383,
2294
+ "loss": 0.7535,
2295
+ "step": 377
2296
+ },
2297
+ {
2298
+ "epoch": 0.62,
2299
+ "learning_rate": 0.00015733389222111592,
2300
+ "loss": 0.7787,
2301
+ "step": 378
2302
+ },
2303
+ {
2304
+ "epoch": 0.62,
2305
+ "learning_rate": 0.0001571199128718076,
2306
+ "loss": 0.7767,
2307
+ "step": 379
2308
+ },
2309
+ {
2310
+ "epoch": 0.63,
2311
+ "learning_rate": 0.00015690554462597522,
2312
+ "loss": 0.7664,
2313
+ "step": 380
2314
+ },
2315
+ {
2316
+ "epoch": 0.63,
2317
+ "learning_rate": 0.00015669078894312848,
2318
+ "loss": 0.784,
2319
+ "step": 381
2320
+ },
2321
+ {
2322
+ "epoch": 0.63,
2323
+ "learning_rate": 0.00015647564728541485,
2324
+ "loss": 0.7815,
2325
+ "step": 382
2326
+ },
2327
+ {
2328
+ "epoch": 0.63,
2329
+ "learning_rate": 0.00015626012111760975,
2330
+ "loss": 0.7598,
2331
+ "step": 383
2332
+ },
2333
+ {
2334
+ "epoch": 0.63,
2335
+ "learning_rate": 0.0001560442119071065,
2336
+ "loss": 0.7546,
2337
+ "step": 384
2338
+ },
2339
+ {
2340
+ "epoch": 0.63,
2341
+ "learning_rate": 0.0001558279211239063,
2342
+ "loss": 0.7328,
2343
+ "step": 385
2344
+ },
2345
+ {
2346
+ "epoch": 0.64,
2347
+ "learning_rate": 0.00015561125024060826,
2348
+ "loss": 0.7731,
2349
+ "step": 386
2350
+ },
2351
+ {
2352
+ "epoch": 0.64,
2353
+ "learning_rate": 0.00015539420073239942,
2354
+ "loss": 0.8272,
2355
+ "step": 387
2356
+ },
2357
+ {
2358
+ "epoch": 0.64,
2359
+ "learning_rate": 0.0001551767740770446,
2360
+ "loss": 0.7338,
2361
+ "step": 388
2362
+ },
2363
+ {
2364
+ "epoch": 0.64,
2365
+ "learning_rate": 0.00015495897175487645,
2366
+ "loss": 0.7247,
2367
+ "step": 389
2368
+ },
2369
+ {
2370
+ "epoch": 0.64,
2371
+ "learning_rate": 0.00015474079524878525,
2372
+ "loss": 0.796,
2373
+ "step": 390
2374
+ },
2375
+ {
2376
+ "epoch": 0.64,
2377
+ "learning_rate": 0.00015452224604420897,
2378
+ "loss": 0.8094,
2379
+ "step": 391
2380
+ },
2381
+ {
2382
+ "epoch": 0.65,
2383
+ "learning_rate": 0.000154303325629123,
2384
+ "loss": 0.768,
2385
+ "step": 392
2386
+ },
2387
+ {
2388
+ "epoch": 0.65,
2389
+ "learning_rate": 0.0001540840354940301,
2390
+ "loss": 0.8052,
2391
+ "step": 393
2392
+ },
2393
+ {
2394
+ "epoch": 0.65,
2395
+ "learning_rate": 0.0001538643771319503,
2396
+ "loss": 0.7602,
2397
+ "step": 394
2398
+ },
2399
+ {
2400
+ "epoch": 0.65,
2401
+ "learning_rate": 0.00015364435203841058,
2402
+ "loss": 0.8005,
2403
+ "step": 395
2404
+ },
2405
+ {
2406
+ "epoch": 0.65,
2407
+ "learning_rate": 0.00015342396171143488,
2408
+ "loss": 0.7279,
2409
+ "step": 396
2410
+ },
2411
+ {
2412
+ "epoch": 0.65,
2413
+ "learning_rate": 0.00015320320765153367,
2414
+ "loss": 0.7832,
2415
+ "step": 397
2416
+ },
2417
+ {
2418
+ "epoch": 0.66,
2419
+ "learning_rate": 0.00015298209136169403,
2420
+ "loss": 0.7498,
2421
+ "step": 398
2422
+ },
2423
+ {
2424
+ "epoch": 0.66,
2425
+ "learning_rate": 0.00015276061434736914,
2426
+ "loss": 0.7825,
2427
+ "step": 399
2428
+ },
2429
+ {
2430
+ "epoch": 0.66,
2431
+ "learning_rate": 0.00015253877811646817,
2432
+ "loss": 0.7973,
2433
+ "step": 400
2434
+ },
2435
+ {
2436
+ "epoch": 0.66,
2437
+ "learning_rate": 0.00015231658417934606,
2438
+ "loss": 0.724,
2439
+ "step": 401
2440
+ },
2441
+ {
2442
+ "epoch": 0.66,
2443
+ "learning_rate": 0.00015209403404879303,
2444
+ "loss": 0.7573,
2445
+ "step": 402
2446
+ },
2447
+ {
2448
+ "epoch": 0.66,
2449
+ "learning_rate": 0.00015187112924002456,
2450
+ "loss": 0.7406,
2451
+ "step": 403
2452
+ },
2453
+ {
2454
+ "epoch": 0.67,
2455
+ "learning_rate": 0.0001516478712706708,
2456
+ "loss": 0.7326,
2457
+ "step": 404
2458
+ },
2459
+ {
2460
+ "epoch": 0.67,
2461
+ "learning_rate": 0.00015142426166076645,
2462
+ "loss": 0.764,
2463
+ "step": 405
2464
+ },
2465
+ {
2466
+ "epoch": 0.67,
2467
+ "learning_rate": 0.00015120030193274027,
2468
+ "loss": 0.6982,
2469
+ "step": 406
2470
+ },
2471
+ {
2472
+ "epoch": 0.67,
2473
+ "learning_rate": 0.00015097599361140487,
2474
+ "loss": 0.7991,
2475
+ "step": 407
2476
+ },
2477
+ {
2478
+ "epoch": 0.67,
2479
+ "learning_rate": 0.00015075133822394613,
2480
+ "loss": 0.7592,
2481
+ "step": 408
2482
+ },
2483
+ {
2484
+ "epoch": 0.67,
2485
+ "learning_rate": 0.00015052633729991294,
2486
+ "loss": 0.7671,
2487
+ "step": 409
2488
+ },
2489
+ {
2490
+ "epoch": 0.68,
2491
+ "learning_rate": 0.00015030099237120674,
2492
+ "loss": 0.7908,
2493
+ "step": 410
2494
+ },
2495
+ {
2496
+ "epoch": 0.68,
2497
+ "learning_rate": 0.00015007530497207117,
2498
+ "loss": 0.7631,
2499
+ "step": 411
2500
+ },
2501
+ {
2502
+ "epoch": 0.68,
2503
+ "learning_rate": 0.00014984927663908137,
2504
+ "loss": 0.73,
2505
+ "step": 412
2506
+ },
2507
+ {
2508
+ "epoch": 0.68,
2509
+ "learning_rate": 0.00014962290891113394,
2510
+ "loss": 0.7648,
2511
+ "step": 413
2512
+ },
2513
+ {
2514
+ "epoch": 0.68,
2515
+ "learning_rate": 0.00014939620332943604,
2516
+ "loss": 0.7777,
2517
+ "step": 414
2518
+ },
2519
+ {
2520
+ "epoch": 0.68,
2521
+ "learning_rate": 0.00014916916143749518,
2522
+ "loss": 0.7699,
2523
+ "step": 415
2524
+ },
2525
+ {
2526
+ "epoch": 0.69,
2527
+ "learning_rate": 0.00014894178478110857,
2528
+ "loss": 0.7755,
2529
+ "step": 416
2530
+ },
2531
+ {
2532
+ "epoch": 0.69,
2533
+ "learning_rate": 0.00014871407490835262,
2534
+ "loss": 0.7863,
2535
+ "step": 417
2536
+ },
2537
+ {
2538
+ "epoch": 0.69,
2539
+ "learning_rate": 0.00014848603336957251,
2540
+ "loss": 0.7413,
2541
+ "step": 418
2542
+ },
2543
+ {
2544
+ "epoch": 0.69,
2545
+ "learning_rate": 0.00014825766171737146,
2546
+ "loss": 0.7917,
2547
+ "step": 419
2548
+ },
2549
+ {
2550
+ "epoch": 0.69,
2551
+ "learning_rate": 0.00014802896150660022,
2552
+ "loss": 0.7522,
2553
+ "step": 420
2554
+ },
2555
+ {
2556
+ "epoch": 0.69,
2557
+ "learning_rate": 0.00014779993429434659,
2558
+ "loss": 0.7361,
2559
+ "step": 421
2560
+ },
2561
+ {
2562
+ "epoch": 0.7,
2563
+ "learning_rate": 0.00014757058163992464,
2564
+ "loss": 0.7713,
2565
+ "step": 422
2566
+ },
2567
+ {
2568
+ "epoch": 0.7,
2569
+ "learning_rate": 0.00014734090510486433,
2570
+ "loss": 0.7535,
2571
+ "step": 423
2572
+ },
2573
+ {
2574
+ "epoch": 0.7,
2575
+ "learning_rate": 0.00014711090625290057,
2576
+ "loss": 0.7542,
2577
+ "step": 424
2578
+ },
2579
+ {
2580
+ "epoch": 0.7,
2581
+ "learning_rate": 0.00014688058664996285,
2582
+ "loss": 0.7781,
2583
+ "step": 425
2584
+ },
2585
+ {
2586
+ "epoch": 0.7,
2587
+ "learning_rate": 0.0001466499478641644,
2588
+ "loss": 0.7643,
2589
+ "step": 426
2590
+ },
2591
+ {
2592
+ "epoch": 0.7,
2593
+ "learning_rate": 0.00014641899146579168,
2594
+ "loss": 0.7743,
2595
+ "step": 427
2596
+ },
2597
+ {
2598
+ "epoch": 0.71,
2599
+ "learning_rate": 0.00014618771902729342,
2600
+ "loss": 0.7716,
2601
+ "step": 428
2602
+ },
2603
+ {
2604
+ "epoch": 0.71,
2605
+ "learning_rate": 0.00014595613212327032,
2606
+ "loss": 0.7737,
2607
+ "step": 429
2608
+ },
2609
+ {
2610
+ "epoch": 0.71,
2611
+ "learning_rate": 0.00014572423233046386,
2612
+ "loss": 0.8296,
2613
+ "step": 430
2614
+ },
2615
+ {
2616
+ "epoch": 0.71,
2617
+ "learning_rate": 0.00014549202122774596,
2618
+ "loss": 0.771,
2619
+ "step": 431
2620
+ },
2621
+ {
2622
+ "epoch": 0.71,
2623
+ "learning_rate": 0.000145259500396108,
2624
+ "loss": 0.7118,
2625
+ "step": 432
2626
+ },
2627
+ {
2628
+ "epoch": 0.71,
2629
+ "learning_rate": 0.00014502667141865015,
2630
+ "loss": 0.7397,
2631
+ "step": 433
2632
+ },
2633
+ {
2634
+ "epoch": 0.71,
2635
+ "learning_rate": 0.00014479353588057052,
2636
+ "loss": 0.7823,
2637
+ "step": 434
2638
+ },
2639
+ {
2640
+ "epoch": 0.72,
2641
+ "learning_rate": 0.00014456009536915448,
2642
+ "loss": 0.7566,
2643
+ "step": 435
2644
+ },
2645
+ {
2646
+ "epoch": 0.72,
2647
+ "learning_rate": 0.00014432635147376376,
2648
+ "loss": 0.7891,
2649
+ "step": 436
2650
+ },
2651
+ {
2652
+ "epoch": 0.72,
2653
+ "learning_rate": 0.00014409230578582566,
2654
+ "loss": 0.7022,
2655
+ "step": 437
2656
+ },
2657
+ {
2658
+ "epoch": 0.72,
2659
+ "learning_rate": 0.00014385795989882221,
2660
+ "loss": 0.75,
2661
+ "step": 438
2662
+ },
2663
+ {
2664
+ "epoch": 0.72,
2665
+ "learning_rate": 0.00014362331540827928,
2666
+ "loss": 0.7041,
2667
+ "step": 439
2668
+ },
2669
+ {
2670
+ "epoch": 0.72,
2671
+ "learning_rate": 0.00014338837391175582,
2672
+ "loss": 0.794,
2673
+ "step": 440
2674
+ },
2675
+ {
2676
+ "epoch": 0.73,
2677
+ "learning_rate": 0.00014315313700883294,
2678
+ "loss": 0.7303,
2679
+ "step": 441
2680
+ },
2681
+ {
2682
+ "epoch": 0.73,
2683
+ "learning_rate": 0.00014291760630110288,
2684
+ "loss": 0.7824,
2685
+ "step": 442
2686
+ },
2687
+ {
2688
+ "epoch": 0.73,
2689
+ "learning_rate": 0.00014268178339215838,
2690
+ "loss": 0.7425,
2691
+ "step": 443
2692
+ },
2693
+ {
2694
+ "epoch": 0.73,
2695
+ "learning_rate": 0.00014244566988758152,
2696
+ "loss": 0.7322,
2697
+ "step": 444
2698
+ },
2699
+ {
2700
+ "epoch": 0.73,
2701
+ "learning_rate": 0.00014220926739493288,
2702
+ "loss": 0.7444,
2703
+ "step": 445
2704
+ },
2705
+ {
2706
+ "epoch": 0.73,
2707
+ "learning_rate": 0.0001419725775237406,
2708
+ "loss": 0.7536,
2709
+ "step": 446
2710
+ },
2711
+ {
2712
+ "epoch": 0.74,
2713
+ "learning_rate": 0.00014173560188548948,
2714
+ "loss": 0.7671,
2715
+ "step": 447
2716
+ },
2717
+ {
2718
+ "epoch": 0.74,
2719
+ "learning_rate": 0.00014149834209360986,
2720
+ "loss": 0.7207,
2721
+ "step": 448
2722
+ },
2723
+ {
2724
+ "epoch": 0.74,
2725
+ "learning_rate": 0.0001412607997634667,
2726
+ "loss": 0.7493,
2727
+ "step": 449
2728
+ },
2729
+ {
2730
+ "epoch": 0.74,
2731
+ "learning_rate": 0.0001410229765123487,
2732
+ "loss": 0.7515,
2733
+ "step": 450
2734
+ },
2735
+ {
2736
+ "epoch": 0.74,
2737
+ "learning_rate": 0.00014078487395945713,
2738
+ "loss": 0.7984,
2739
+ "step": 451
2740
+ },
2741
+ {
2742
+ "epoch": 0.74,
2743
+ "learning_rate": 0.00014054649372589482,
2744
+ "loss": 0.7558,
2745
+ "step": 452
2746
+ },
2747
+ {
2748
+ "epoch": 0.75,
2749
+ "learning_rate": 0.00014030783743465528,
2750
+ "loss": 0.7545,
2751
+ "step": 453
2752
+ },
2753
+ {
2754
+ "epoch": 0.75,
2755
+ "learning_rate": 0.00014006890671061143,
2756
+ "loss": 0.7613,
2757
+ "step": 454
2758
+ },
2759
+ {
2760
+ "epoch": 0.75,
2761
+ "learning_rate": 0.0001398297031805047,
2762
+ "loss": 0.7827,
2763
+ "step": 455
2764
+ },
2765
+ {
2766
+ "epoch": 0.75,
2767
+ "learning_rate": 0.00013959022847293391,
2768
+ "loss": 0.7348,
2769
+ "step": 456
2770
+ },
2771
+ {
2772
+ "epoch": 0.75,
2773
+ "eval_loss": 0.7748340368270874,
2774
+ "eval_runtime": 85.3906,
2775
+ "eval_samples_per_second": 4.029,
2776
+ "eval_steps_per_second": 2.014,
2777
+ "step": 456
2778
+ },
2779
+ {
2780
+ "epoch": 0.75,
2781
+ "learning_rate": 0.0001393504842183441,
2782
+ "loss": 0.8026,
2783
+ "step": 457
2784
+ },
2785
+ {
2786
+ "epoch": 0.75,
2787
+ "learning_rate": 0.0001391104720490156,
2788
+ "loss": 0.7807,
2789
+ "step": 458
2790
+ },
2791
+ {
2792
+ "epoch": 0.76,
2793
+ "learning_rate": 0.00013887019359905275,
2794
+ "loss": 0.7578,
2795
+ "step": 459
2796
+ },
2797
+ {
2798
+ "epoch": 0.76,
2799
+ "learning_rate": 0.0001386296505043728,
2800
+ "loss": 0.7462,
2801
+ "step": 460
2802
+ },
2803
+ {
2804
+ "epoch": 0.76,
2805
+ "learning_rate": 0.00013838884440269496,
2806
+ "loss": 0.7644,
2807
+ "step": 461
2808
+ },
2809
+ {
2810
+ "epoch": 0.76,
2811
+ "learning_rate": 0.000138147776933529,
2812
+ "loss": 0.7897,
2813
+ "step": 462
2814
+ },
2815
+ {
2816
+ "epoch": 0.76,
2817
+ "learning_rate": 0.0001379064497381641,
2818
+ "loss": 0.7399,
2819
+ "step": 463
2820
+ },
2821
+ {
2822
+ "epoch": 0.76,
2823
+ "learning_rate": 0.00013766486445965795,
2824
+ "loss": 0.767,
2825
+ "step": 464
2826
+ },
2827
+ {
2828
+ "epoch": 0.77,
2829
+ "learning_rate": 0.00013742302274282533,
2830
+ "loss": 0.7733,
2831
+ "step": 465
2832
+ },
2833
+ {
2834
+ "epoch": 0.77,
2835
+ "learning_rate": 0.00013718092623422686,
2836
+ "loss": 0.7551,
2837
+ "step": 466
2838
+ },
2839
+ {
2840
+ "epoch": 0.77,
2841
+ "learning_rate": 0.000136938576582158,
2842
+ "loss": 0.7815,
2843
+ "step": 467
2844
+ },
2845
+ {
2846
+ "epoch": 0.77,
2847
+ "learning_rate": 0.00013669597543663762,
2848
+ "loss": 0.74,
2849
+ "step": 468
2850
+ },
2851
+ {
2852
+ "epoch": 0.77,
2853
+ "learning_rate": 0.000136453124449397,
2854
+ "loss": 0.7374,
2855
+ "step": 469
2856
+ },
2857
+ {
2858
+ "epoch": 0.77,
2859
+ "learning_rate": 0.00013621002527386834,
2860
+ "loss": 0.741,
2861
+ "step": 470
2862
+ },
2863
+ {
2864
+ "epoch": 0.78,
2865
+ "learning_rate": 0.0001359666795651736,
2866
+ "loss": 0.7938,
2867
+ "step": 471
2868
+ },
2869
+ {
2870
+ "epoch": 0.78,
2871
+ "learning_rate": 0.0001357230889801133,
2872
+ "loss": 0.7252,
2873
+ "step": 472
2874
+ },
2875
+ {
2876
+ "epoch": 0.78,
2877
+ "learning_rate": 0.00013547925517715519,
2878
+ "loss": 0.7318,
2879
+ "step": 473
2880
+ },
2881
+ {
2882
+ "epoch": 0.78,
2883
+ "learning_rate": 0.00013523517981642286,
2884
+ "loss": 0.741,
2885
+ "step": 474
2886
+ },
2887
+ {
2888
+ "epoch": 0.78,
2889
+ "learning_rate": 0.00013499086455968467,
2890
+ "loss": 0.7595,
2891
+ "step": 475
2892
+ },
2893
+ {
2894
+ "epoch": 0.78,
2895
+ "learning_rate": 0.0001347463110703422,
2896
+ "loss": 0.7767,
2897
+ "step": 476
2898
+ },
2899
+ {
2900
+ "epoch": 0.79,
2901
+ "learning_rate": 0.00013450152101341896,
2902
+ "loss": 0.7698,
2903
+ "step": 477
2904
+ },
2905
+ {
2906
+ "epoch": 0.79,
2907
+ "learning_rate": 0.00013425649605554928,
2908
+ "loss": 0.7661,
2909
+ "step": 478
2910
+ },
2911
+ {
2912
+ "epoch": 0.79,
2913
+ "learning_rate": 0.00013401123786496664,
2914
+ "loss": 0.7558,
2915
+ "step": 479
2916
+ },
2917
+ {
2918
+ "epoch": 0.79,
2919
+ "learning_rate": 0.00013376574811149253,
2920
+ "loss": 0.745,
2921
+ "step": 480
2922
+ },
2923
+ {
2924
+ "epoch": 0.79,
2925
+ "learning_rate": 0.00013352002846652504,
2926
+ "loss": 0.7501,
2927
+ "step": 481
2928
+ },
2929
+ {
2930
+ "epoch": 0.79,
2931
+ "learning_rate": 0.00013327408060302738,
2932
+ "loss": 0.7399,
2933
+ "step": 482
2934
+ },
2935
+ {
2936
+ "epoch": 0.8,
2937
+ "learning_rate": 0.00013302790619551674,
2938
+ "loss": 0.7038,
2939
+ "step": 483
2940
+ },
2941
+ {
2942
+ "epoch": 0.8,
2943
+ "learning_rate": 0.00013278150692005243,
2944
+ "loss": 0.7526,
2945
+ "step": 484
2946
+ },
2947
+ {
2948
+ "epoch": 0.8,
2949
+ "learning_rate": 0.00013253488445422507,
2950
+ "loss": 0.7493,
2951
+ "step": 485
2952
+ },
2953
+ {
2954
+ "epoch": 0.8,
2955
+ "learning_rate": 0.00013228804047714463,
2956
+ "loss": 0.7053,
2957
+ "step": 486
2958
+ },
2959
+ {
2960
+ "epoch": 0.8,
2961
+ "learning_rate": 0.00013204097666942932,
2962
+ "loss": 0.7826,
2963
+ "step": 487
2964
+ },
2965
+ {
2966
+ "epoch": 0.8,
2967
+ "learning_rate": 0.00013179369471319404,
2968
+ "loss": 0.7719,
2969
+ "step": 488
2970
+ },
2971
+ {
2972
+ "epoch": 0.81,
2973
+ "learning_rate": 0.00013154619629203893,
2974
+ "loss": 0.7668,
2975
+ "step": 489
2976
+ },
2977
+ {
2978
+ "epoch": 0.81,
2979
+ "learning_rate": 0.0001312984830910379,
2980
+ "loss": 0.762,
2981
+ "step": 490
2982
+ },
2983
+ {
2984
+ "epoch": 0.81,
2985
+ "learning_rate": 0.0001310505567967272,
2986
+ "loss": 0.7346,
2987
+ "step": 491
2988
+ },
2989
+ {
2990
+ "epoch": 0.81,
2991
+ "learning_rate": 0.00013080241909709387,
2992
+ "loss": 0.7595,
2993
+ "step": 492
2994
+ },
2995
+ {
2996
+ "epoch": 0.81,
2997
+ "learning_rate": 0.00013055407168156437,
2998
+ "loss": 0.7422,
2999
+ "step": 493
3000
+ },
3001
+ {
3002
+ "epoch": 0.81,
3003
+ "learning_rate": 0.00013030551624099287,
3004
+ "loss": 0.7342,
3005
+ "step": 494
3006
+ },
3007
+ {
3008
+ "epoch": 0.82,
3009
+ "learning_rate": 0.00013005675446764998,
3010
+ "loss": 0.7401,
3011
+ "step": 495
3012
+ },
3013
+ {
3014
+ "epoch": 0.82,
3015
+ "learning_rate": 0.000129807788055211,
3016
+ "loss": 0.7577,
3017
+ "step": 496
3018
+ },
3019
+ {
3020
+ "epoch": 0.82,
3021
+ "learning_rate": 0.0001295586186987446,
3022
+ "loss": 0.7662,
3023
+ "step": 497
3024
+ },
3025
+ {
3026
+ "epoch": 0.82,
3027
+ "learning_rate": 0.00012930924809470115,
3028
+ "loss": 0.7131,
3029
+ "step": 498
3030
+ },
3031
+ {
3032
+ "epoch": 0.82,
3033
+ "learning_rate": 0.00012905967794090114,
3034
+ "loss": 0.7217,
3035
+ "step": 499
3036
+ },
3037
+ {
3038
+ "epoch": 0.82,
3039
+ "learning_rate": 0.00012880990993652377,
3040
+ "loss": 0.7616,
3041
+ "step": 500
3042
+ },
3043
+ {
3044
+ "epoch": 0.83,
3045
+ "learning_rate": 0.00012855994578209526,
3046
+ "loss": 0.7784,
3047
+ "step": 501
3048
+ },
3049
+ {
3050
+ "epoch": 0.83,
3051
+ "learning_rate": 0.00012830978717947718,
3052
+ "loss": 0.7677,
3053
+ "step": 502
3054
+ },
3055
+ {
3056
+ "epoch": 0.83,
3057
+ "learning_rate": 0.00012805943583185525,
3058
+ "loss": 0.7629,
3059
+ "step": 503
3060
+ },
3061
+ {
3062
+ "epoch": 0.83,
3063
+ "learning_rate": 0.00012780889344372718,
3064
+ "loss": 0.7648,
3065
+ "step": 504
3066
+ },
3067
+ {
3068
+ "epoch": 0.83,
3069
+ "learning_rate": 0.00012755816172089164,
3070
+ "loss": 0.785,
3071
+ "step": 505
3072
+ },
3073
+ {
3074
+ "epoch": 0.83,
3075
+ "learning_rate": 0.00012730724237043615,
3076
+ "loss": 0.7422,
3077
+ "step": 506
3078
+ },
3079
+ {
3080
+ "epoch": 0.84,
3081
+ "learning_rate": 0.00012705613710072575,
3082
+ "loss": 0.7862,
3083
+ "step": 507
3084
+ },
3085
+ {
3086
+ "epoch": 0.84,
3087
+ "learning_rate": 0.0001268048476213914,
3088
+ "loss": 0.7736,
3089
+ "step": 508
3090
+ },
3091
+ {
3092
+ "epoch": 0.84,
3093
+ "learning_rate": 0.00012655337564331805,
3094
+ "loss": 0.7511,
3095
+ "step": 509
3096
+ },
3097
+ {
3098
+ "epoch": 0.84,
3099
+ "learning_rate": 0.0001263017228786334,
3100
+ "loss": 0.7707,
3101
+ "step": 510
3102
+ },
3103
+ {
3104
+ "epoch": 0.84,
3105
+ "learning_rate": 0.0001260498910406958,
3106
+ "loss": 0.7744,
3107
+ "step": 511
3108
+ },
3109
+ {
3110
+ "epoch": 0.84,
3111
+ "learning_rate": 0.00012579788184408295,
3112
+ "loss": 0.7391,
3113
+ "step": 512
3114
+ },
3115
+ {
3116
+ "epoch": 0.85,
3117
+ "learning_rate": 0.00012554569700458002,
3118
+ "loss": 0.6762,
3119
+ "step": 513
3120
+ },
3121
+ {
3122
+ "epoch": 0.85,
3123
+ "learning_rate": 0.00012529333823916807,
3124
+ "loss": 0.7606,
3125
+ "step": 514
3126
+ },
3127
+ {
3128
+ "epoch": 0.85,
3129
+ "learning_rate": 0.00012504080726601232,
3130
+ "loss": 0.7928,
3131
+ "step": 515
3132
+ },
3133
+ {
3134
+ "epoch": 0.85,
3135
+ "learning_rate": 0.0001247881058044504,
3136
+ "loss": 0.7404,
3137
+ "step": 516
3138
+ },
3139
+ {
3140
+ "epoch": 0.85,
3141
+ "learning_rate": 0.00012453523557498075,
3142
+ "loss": 0.7493,
3143
+ "step": 517
3144
+ },
3145
+ {
3146
+ "epoch": 0.85,
3147
+ "learning_rate": 0.00012428219829925083,
3148
+ "loss": 0.7582,
3149
+ "step": 518
3150
+ },
3151
+ {
3152
+ "epoch": 0.86,
3153
+ "learning_rate": 0.00012402899570004543,
3154
+ "loss": 0.7012,
3155
+ "step": 519
3156
+ },
3157
+ {
3158
+ "epoch": 0.86,
3159
+ "learning_rate": 0.00012377562950127493,
3160
+ "loss": 0.7698,
3161
+ "step": 520
3162
+ },
3163
+ {
3164
+ "epoch": 0.86,
3165
+ "learning_rate": 0.0001235221014279636,
3166
+ "loss": 0.7408,
3167
+ "step": 521
3168
+ },
3169
+ {
3170
+ "epoch": 0.86,
3171
+ "learning_rate": 0.00012326841320623767,
3172
+ "loss": 0.7724,
3173
+ "step": 522
3174
+ },
3175
+ {
3176
+ "epoch": 0.86,
3177
+ "learning_rate": 0.00012301456656331402,
3178
+ "loss": 0.7512,
3179
+ "step": 523
3180
+ },
3181
+ {
3182
+ "epoch": 0.86,
3183
+ "learning_rate": 0.00012276056322748778,
3184
+ "loss": 0.7509,
3185
+ "step": 524
3186
+ },
3187
+ {
3188
+ "epoch": 0.86,
3189
+ "learning_rate": 0.0001225064049281212,
3190
+ "loss": 0.7561,
3191
+ "step": 525
3192
+ },
3193
+ {
3194
+ "epoch": 0.87,
3195
+ "learning_rate": 0.00012225209339563145,
3196
+ "loss": 0.7758,
3197
+ "step": 526
3198
+ },
3199
+ {
3200
+ "epoch": 0.87,
3201
+ "learning_rate": 0.00012199763036147895,
3202
+ "loss": 0.8028,
3203
+ "step": 527
3204
+ },
3205
+ {
3206
+ "epoch": 0.87,
3207
+ "learning_rate": 0.00012174301755815571,
3208
+ "loss": 0.7113,
3209
+ "step": 528
3210
+ },
3211
+ {
3212
+ "epoch": 0.87,
3213
+ "learning_rate": 0.00012148825671917334,
3214
+ "loss": 0.7941,
3215
+ "step": 529
3216
+ },
3217
+ {
3218
+ "epoch": 0.87,
3219
+ "learning_rate": 0.0001212333495790514,
3220
+ "loss": 0.7639,
3221
+ "step": 530
3222
+ },
3223
+ {
3224
+ "epoch": 0.87,
3225
+ "learning_rate": 0.00012097829787330544,
3226
+ "loss": 0.7481,
3227
+ "step": 531
3228
+ },
3229
+ {
3230
+ "epoch": 0.88,
3231
+ "learning_rate": 0.00012072310333843544,
3232
+ "loss": 0.7732,
3233
+ "step": 532
3234
+ },
3235
+ {
3236
+ "epoch": 0.88,
3237
+ "learning_rate": 0.00012046776771191366,
3238
+ "loss": 0.7493,
3239
+ "step": 533
3240
+ },
3241
+ {
3242
+ "epoch": 0.88,
3243
+ "learning_rate": 0.00012021229273217302,
3244
+ "loss": 0.7359,
3245
+ "step": 534
3246
+ },
3247
+ {
3248
+ "epoch": 0.88,
3249
+ "learning_rate": 0.00011995668013859529,
3250
+ "loss": 0.7319,
3251
+ "step": 535
3252
+ },
3253
+ {
3254
+ "epoch": 0.88,
3255
+ "learning_rate": 0.00011970093167149905,
3256
+ "loss": 0.7671,
3257
+ "step": 536
3258
+ },
3259
+ {
3260
+ "epoch": 0.88,
3261
+ "learning_rate": 0.00011944504907212804,
3262
+ "loss": 0.7514,
3263
+ "step": 537
3264
+ },
3265
+ {
3266
+ "epoch": 0.89,
3267
+ "learning_rate": 0.00011918903408263924,
3268
+ "loss": 0.7095,
3269
+ "step": 538
3270
+ },
3271
+ {
3272
+ "epoch": 0.89,
3273
+ "learning_rate": 0.00011893288844609094,
3274
+ "loss": 0.7625,
3275
+ "step": 539
3276
+ },
3277
+ {
3278
+ "epoch": 0.89,
3279
+ "learning_rate": 0.000118676613906431,
3280
+ "loss": 0.7346,
3281
+ "step": 540
3282
+ },
3283
+ {
3284
+ "epoch": 0.89,
3285
+ "learning_rate": 0.00011842021220848486,
3286
+ "loss": 0.7206,
3287
+ "step": 541
3288
+ },
3289
+ {
3290
+ "epoch": 0.89,
3291
+ "learning_rate": 0.00011816368509794364,
3292
+ "loss": 0.7408,
3293
+ "step": 542
3294
+ },
3295
+ {
3296
+ "epoch": 0.89,
3297
+ "learning_rate": 0.00011790703432135253,
3298
+ "loss": 0.7213,
3299
+ "step": 543
3300
+ },
3301
+ {
3302
+ "epoch": 0.9,
3303
+ "learning_rate": 0.00011765026162609847,
3304
+ "loss": 0.7476,
3305
+ "step": 544
3306
+ },
3307
+ {
3308
+ "epoch": 0.9,
3309
+ "learning_rate": 0.00011739336876039859,
3310
+ "loss": 0.7787,
3311
+ "step": 545
3312
+ },
3313
+ {
3314
+ "epoch": 0.9,
3315
+ "learning_rate": 0.00011713635747328818,
3316
+ "loss": 0.7646,
3317
+ "step": 546
3318
+ },
3319
+ {
3320
+ "epoch": 0.9,
3321
+ "learning_rate": 0.00011687922951460872,
3322
+ "loss": 0.6935,
3323
+ "step": 547
3324
+ },
3325
+ {
3326
+ "epoch": 0.9,
3327
+ "learning_rate": 0.00011662198663499619,
3328
+ "loss": 0.7403,
3329
+ "step": 548
3330
+ },
3331
+ {
3332
+ "epoch": 0.9,
3333
+ "learning_rate": 0.00011636463058586881,
3334
+ "loss": 0.804,
3335
+ "step": 549
3336
+ },
3337
+ {
3338
+ "epoch": 0.91,
3339
+ "learning_rate": 0.0001161071631194155,
3340
+ "loss": 0.7442,
3341
+ "step": 550
3342
+ },
3343
+ {
3344
+ "epoch": 0.91,
3345
+ "learning_rate": 0.00011584958598858359,
3346
+ "loss": 0.76,
3347
+ "step": 551
3348
+ },
3349
+ {
3350
+ "epoch": 0.91,
3351
+ "learning_rate": 0.00011559190094706714,
3352
+ "loss": 0.7403,
3353
+ "step": 552
3354
+ },
3355
+ {
3356
+ "epoch": 0.91,
3357
+ "learning_rate": 0.0001153341097492949,
3358
+ "loss": 0.7514,
3359
+ "step": 553
3360
+ },
3361
+ {
3362
+ "epoch": 0.91,
3363
+ "learning_rate": 0.00011507621415041837,
3364
+ "loss": 0.7438,
3365
+ "step": 554
3366
+ },
3367
+ {
3368
+ "epoch": 0.91,
3369
+ "learning_rate": 0.00011481821590629985,
3370
+ "loss": 0.725,
3371
+ "step": 555
3372
+ },
3373
+ {
3374
+ "epoch": 0.92,
3375
+ "learning_rate": 0.00011456011677350051,
3376
+ "loss": 0.6918,
3377
+ "step": 556
3378
+ },
3379
+ {
3380
+ "epoch": 0.92,
3381
+ "learning_rate": 0.00011430191850926837,
3382
+ "loss": 0.767,
3383
+ "step": 557
3384
+ },
3385
+ {
3386
+ "epoch": 0.92,
3387
+ "learning_rate": 0.00011404362287152646,
3388
+ "loss": 0.7356,
3389
+ "step": 558
3390
+ },
3391
+ {
3392
+ "epoch": 0.92,
3393
+ "learning_rate": 0.00011378523161886066,
3394
+ "loss": 0.7281,
3395
+ "step": 559
3396
+ },
3397
+ {
3398
+ "epoch": 0.92,
3399
+ "learning_rate": 0.00011352674651050796,
3400
+ "loss": 0.7428,
3401
+ "step": 560
3402
+ },
3403
+ {
3404
+ "epoch": 0.92,
3405
+ "learning_rate": 0.00011326816930634427,
3406
+ "loss": 0.7005,
3407
+ "step": 561
3408
+ },
3409
+ {
3410
+ "epoch": 0.93,
3411
+ "learning_rate": 0.00011300950176687255,
3412
+ "loss": 0.7109,
3413
+ "step": 562
3414
+ },
3415
+ {
3416
+ "epoch": 0.93,
3417
+ "learning_rate": 0.0001127507456532108,
3418
+ "loss": 0.7378,
3419
+ "step": 563
3420
+ },
3421
+ {
3422
+ "epoch": 0.93,
3423
+ "learning_rate": 0.00011249190272708008,
3424
+ "loss": 0.7413,
3425
+ "step": 564
3426
+ },
3427
+ {
3428
+ "epoch": 0.93,
3429
+ "learning_rate": 0.00011223297475079251,
3430
+ "loss": 0.7741,
3431
+ "step": 565
3432
+ },
3433
+ {
3434
+ "epoch": 0.93,
3435
+ "learning_rate": 0.00011197396348723923,
3436
+ "loss": 0.7547,
3437
+ "step": 566
3438
+ },
3439
+ {
3440
+ "epoch": 0.93,
3441
+ "learning_rate": 0.00011171487069987851,
3442
+ "loss": 0.7478,
3443
+ "step": 567
3444
+ },
3445
+ {
3446
+ "epoch": 0.94,
3447
+ "learning_rate": 0.0001114556981527236,
3448
+ "loss": 0.7138,
3449
+ "step": 568
3450
+ },
3451
+ {
3452
+ "epoch": 0.94,
3453
+ "learning_rate": 0.00011119644761033078,
3454
+ "loss": 0.76,
3455
+ "step": 569
3456
+ },
3457
+ {
3458
+ "epoch": 0.94,
3459
+ "learning_rate": 0.00011093712083778746,
3460
+ "loss": 0.771,
3461
+ "step": 570
3462
+ },
3463
+ {
3464
+ "epoch": 0.94,
3465
+ "learning_rate": 0.00011067771960069991,
3466
+ "loss": 0.765,
3467
+ "step": 571
3468
+ },
3469
+ {
3470
+ "epoch": 0.94,
3471
+ "learning_rate": 0.00011041824566518146,
3472
+ "loss": 0.7886,
3473
+ "step": 572
3474
+ },
3475
+ {
3476
+ "epoch": 0.94,
3477
+ "learning_rate": 0.00011015870079784048,
3478
+ "loss": 0.7134,
3479
+ "step": 573
3480
+ },
3481
+ {
3482
+ "epoch": 0.95,
3483
+ "learning_rate": 0.00010989908676576807,
3484
+ "loss": 0.7173,
3485
+ "step": 574
3486
+ },
3487
+ {
3488
+ "epoch": 0.95,
3489
+ "learning_rate": 0.00010963940533652648,
3490
+ "loss": 0.7522,
3491
+ "step": 575
3492
+ },
3493
+ {
3494
+ "epoch": 0.95,
3495
+ "learning_rate": 0.00010937965827813661,
3496
+ "loss": 0.7523,
3497
+ "step": 576
3498
+ },
3499
+ {
3500
+ "epoch": 0.95,
3501
+ "learning_rate": 0.00010911984735906635,
3502
+ "loss": 0.7446,
3503
+ "step": 577
3504
+ },
3505
+ {
3506
+ "epoch": 0.95,
3507
+ "learning_rate": 0.00010885997434821831,
3508
+ "loss": 0.7226,
3509
+ "step": 578
3510
+ },
3511
+ {
3512
+ "epoch": 0.95,
3513
+ "learning_rate": 0.00010860004101491779,
3514
+ "loss": 0.7181,
3515
+ "step": 579
3516
+ },
3517
+ {
3518
+ "epoch": 0.96,
3519
+ "learning_rate": 0.00010834004912890092,
3520
+ "loss": 0.7689,
3521
+ "step": 580
3522
+ },
3523
+ {
3524
+ "epoch": 0.96,
3525
+ "learning_rate": 0.0001080800004603024,
3526
+ "loss": 0.7329,
3527
+ "step": 581
3528
+ },
3529
+ {
3530
+ "epoch": 0.96,
3531
+ "learning_rate": 0.00010781989677964355,
3532
+ "loss": 0.7067,
3533
+ "step": 582
3534
+ },
3535
+ {
3536
+ "epoch": 0.96,
3537
+ "learning_rate": 0.00010755973985782022,
3538
+ "loss": 0.7191,
3539
+ "step": 583
3540
+ },
3541
+ {
3542
+ "epoch": 0.96,
3543
+ "learning_rate": 0.00010729953146609076,
3544
+ "loss": 0.716,
3545
+ "step": 584
3546
+ },
3547
+ {
3548
+ "epoch": 0.96,
3549
+ "learning_rate": 0.00010703927337606396,
3550
+ "loss": 0.71,
3551
+ "step": 585
3552
+ },
3553
+ {
3554
+ "epoch": 0.97,
3555
+ "learning_rate": 0.00010677896735968693,
3556
+ "loss": 0.7327,
3557
+ "step": 586
3558
+ },
3559
+ {
3560
+ "epoch": 0.97,
3561
+ "learning_rate": 0.00010651861518923319,
3562
+ "loss": 0.7322,
3563
+ "step": 587
3564
+ },
3565
+ {
3566
+ "epoch": 0.97,
3567
+ "learning_rate": 0.00010625821863729036,
3568
+ "loss": 0.7429,
3569
+ "step": 588
3570
+ },
3571
+ {
3572
+ "epoch": 0.97,
3573
+ "learning_rate": 0.00010599777947674829,
3574
+ "loss": 0.7499,
3575
+ "step": 589
3576
+ },
3577
+ {
3578
+ "epoch": 0.97,
3579
+ "learning_rate": 0.00010573729948078699,
3580
+ "loss": 0.7391,
3581
+ "step": 590
3582
+ },
3583
+ {
3584
+ "epoch": 0.97,
3585
+ "learning_rate": 0.00010547678042286436,
3586
+ "loss": 0.7499,
3587
+ "step": 591
3588
+ },
3589
+ {
3590
+ "epoch": 0.98,
3591
+ "learning_rate": 0.00010521622407670439,
3592
+ "loss": 0.7398,
3593
+ "step": 592
3594
+ },
3595
+ {
3596
+ "epoch": 0.98,
3597
+ "learning_rate": 0.00010495563221628486,
3598
+ "loss": 0.7206,
3599
+ "step": 593
3600
+ },
3601
+ {
3602
+ "epoch": 0.98,
3603
+ "learning_rate": 0.00010469500661582536,
3604
+ "loss": 0.7797,
3605
+ "step": 594
3606
+ },
3607
+ {
3608
+ "epoch": 0.98,
3609
+ "learning_rate": 0.00010443434904977518,
3610
+ "loss": 0.7936,
3611
+ "step": 595
3612
+ },
3613
+ {
3614
+ "epoch": 0.98,
3615
+ "learning_rate": 0.00010417366129280133,
3616
+ "loss": 0.7512,
3617
+ "step": 596
3618
+ },
3619
+ {
3620
+ "epoch": 0.98,
3621
+ "learning_rate": 0.00010391294511977623,
3622
+ "loss": 0.7616,
3623
+ "step": 597
3624
+ },
3625
+ {
3626
+ "epoch": 0.99,
3627
+ "learning_rate": 0.0001036522023057659,
3628
+ "loss": 0.7767,
3629
+ "step": 598
3630
+ },
3631
+ {
3632
+ "epoch": 0.99,
3633
+ "learning_rate": 0.00010339143462601768,
3634
+ "loss": 0.7676,
3635
+ "step": 599
3636
+ },
3637
+ {
3638
+ "epoch": 0.99,
3639
+ "learning_rate": 0.00010313064385594822,
3640
+ "loss": 0.7357,
3641
+ "step": 600
3642
+ },
3643
+ {
3644
+ "epoch": 0.99,
3645
+ "learning_rate": 0.00010286983177113135,
3646
+ "loss": 0.7563,
3647
+ "step": 601
3648
+ },
3649
+ {
3650
+ "epoch": 0.99,
3651
+ "learning_rate": 0.0001026090001472861,
3652
+ "loss": 0.7228,
3653
+ "step": 602
3654
+ },
3655
+ {
3656
+ "epoch": 0.99,
3657
+ "learning_rate": 0.00010234815076026442,
3658
+ "loss": 0.7603,
3659
+ "step": 603
3660
+ },
3661
+ {
3662
+ "epoch": 1.0,
3663
+ "learning_rate": 0.00010208728538603929,
3664
+ "loss": 0.7354,
3665
+ "step": 604
3666
+ },
3667
+ {
3668
+ "epoch": 1.0,
3669
+ "learning_rate": 0.0001018264058006925,
3670
+ "loss": 0.7657,
3671
+ "step": 605
3672
+ },
3673
+ {
3674
+ "epoch": 1.0,
3675
+ "learning_rate": 0.00010156551378040258,
3676
+ "loss": 0.7103,
3677
+ "step": 606
3678
+ },
3679
+ {
3680
+ "epoch": 1.0,
3681
+ "learning_rate": 0.00010130461110143277,
3682
+ "loss": 0.7663,
3683
+ "step": 607
3684
+ }
3685
+ ],
3686
+ "logging_steps": 1,
3687
+ "max_steps": 1214,
3688
+ "num_input_tokens_seen": 0,
3689
+ "num_train_epochs": 2,
3690
+ "save_steps": 607,
3691
+ "total_flos": 3.9838190722351104e+17,
3692
+ "train_batch_size": 2,
3693
+ "trial_name": null,
3694
+ "trial_params": null
3695
+ }
checkpoint-607/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0b7fb8249d23a78d379d117f7117168f666a5d46ad4a572405b96de4612aa1d7
3
+ size 4731
config.json ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "NousResearch/Llama-2-13b-hf",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "silu",
11
+ "hidden_size": 5120,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 13824,
14
+ "max_position_embeddings": 4096,
15
+ "model_type": "llama",
16
+ "num_attention_heads": 40,
17
+ "num_hidden_layers": 40,
18
+ "num_key_value_heads": 40,
19
+ "pad_token_id": 0,
20
+ "pretraining_tp": 1,
21
+ "quantization_config": {
22
+ "bnb_4bit_compute_dtype": "float32",
23
+ "bnb_4bit_quant_type": "fp4",
24
+ "bnb_4bit_use_double_quant": false,
25
+ "llm_int8_enable_fp32_cpu_offload": false,
26
+ "llm_int8_has_fp16_weight": false,
27
+ "llm_int8_skip_modules": null,
28
+ "llm_int8_threshold": 6.0,
29
+ "load_in_4bit": false,
30
+ "load_in_8bit": true,
31
+ "quant_method": "bitsandbytes"
32
+ },
33
+ "rms_norm_eps": 1e-05,
34
+ "rope_scaling": null,
35
+ "rope_theta": 10000.0,
36
+ "tie_word_embeddings": false,
37
+ "torch_dtype": "float16",
38
+ "transformers_version": "4.36.2",
39
+ "use_cache": false,
40
+ "vocab_size": 32000
41
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": {
17
+ "content": "<unk>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "unk_token": {
24
+ "content": "<unk>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ }
30
+ }
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 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "<unk>",
36
+ "sp_model_kwargs": {},
37
+ "spaces_between_special_tokens": false,
38
+ "tokenizer_class": "LlamaTokenizer",
39
+ "trust_remote_code": false,
40
+ "unk_token": "<unk>",
41
+ "use_default_system_prompt": false,
42
+ "use_fast": true
43
+ }