yalhessi commited on
Commit
25e14c5
·
verified ·
1 Parent(s): 01106c8

Training in progress, epoch 12, checkpoint

Browse files
checkpoint-186828/README.md ADDED
@@ -0,0 +1,202 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: deepseek-ai/deepseek-coder-6.7b-base
3
+ library_name: peft
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
+ ### Framework versions
201
+
202
+ - PEFT 0.14.0
checkpoint-186828/adapter_config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "deepseek-ai/deepseek-coder-6.7b-base",
5
+ "bias": "none",
6
+ "eva_config": null,
7
+ "exclude_modules": null,
8
+ "fan_in_fan_out": false,
9
+ "inference_mode": true,
10
+ "init_lora_weights": true,
11
+ "layer_replication": null,
12
+ "layers_pattern": null,
13
+ "layers_to_transform": null,
14
+ "loftq_config": {},
15
+ "lora_alpha": 32,
16
+ "lora_bias": false,
17
+ "lora_dropout": 0.05,
18
+ "megatron_config": null,
19
+ "megatron_core": "megatron.core",
20
+ "modules_to_save": null,
21
+ "peft_type": "LORA",
22
+ "r": 8,
23
+ "rank_pattern": {},
24
+ "revision": null,
25
+ "target_modules": [
26
+ "v_proj",
27
+ "q_proj"
28
+ ],
29
+ "task_type": "CAUSAL_LM",
30
+ "use_dora": false,
31
+ "use_rslora": false
32
+ }
checkpoint-186828/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e58eeafa2d1fb1b825bf6bcac49a849c591d6ce86e6119b35cbcbf29b709346d
3
+ size 541459256
checkpoint-186828/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fa36a6dc80cb79b86b05684d6a43cfef967b9a2590f4a17c65137ec55d8f3675
3
+ size 33662074
checkpoint-186828/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8036d1cdc526b543594008d4c2319fe3040157da3b9ff9846ebb06361f8431b3
3
+ size 15024
checkpoint-186828/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:efff1dfa216abbba6cf577f2c9f74430d7db25b15c773855e037c98c9aea759b
3
+ size 15024
checkpoint-186828/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26b936db45db9454b8ab655a1b03e9280fa6d4b355d86cd3775c49a786cc3e69
3
+ size 15024
checkpoint-186828/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c91a62c7b350e7c8637561d4c5e6eebc699bdf3400925e3516d268af7f910dbc
3
+ size 15024
checkpoint-186828/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac9232235bf4e910d89bceeda37e905b114a386072ebb3e13f4532e7e24c832f
3
+ size 1064
checkpoint-186828/trainer_state.json ADDED
@@ -0,0 +1,3116 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 12.0,
5
+ "eval_steps": 3114,
6
+ "global_step": 186828,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.03211510052026463,
13
+ "grad_norm": 0.6881807446479797,
14
+ "learning_rate": 0.00039893377866272726,
15
+ "loss": 0.3182,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 0.06423020104052926,
20
+ "grad_norm": 0.40920817852020264,
21
+ "learning_rate": 0.00039786327531205173,
22
+ "loss": 0.2226,
23
+ "step": 1000
24
+ },
25
+ {
26
+ "epoch": 0.09634530156079389,
27
+ "grad_norm": 0.38458916544914246,
28
+ "learning_rate": 0.00039679277196137626,
29
+ "loss": 0.2056,
30
+ "step": 1500
31
+ },
32
+ {
33
+ "epoch": 0.12846040208105852,
34
+ "grad_norm": 0.3895084857940674,
35
+ "learning_rate": 0.0003957222686107008,
36
+ "loss": 0.1921,
37
+ "step": 2000
38
+ },
39
+ {
40
+ "epoch": 0.16057550260132314,
41
+ "grad_norm": 1.2132288217544556,
42
+ "learning_rate": 0.00039465176526002526,
43
+ "loss": 0.1828,
44
+ "step": 2500
45
+ },
46
+ {
47
+ "epoch": 0.19269060312158778,
48
+ "grad_norm": 0.354610413312912,
49
+ "learning_rate": 0.0003935812619093498,
50
+ "loss": 0.1813,
51
+ "step": 3000
52
+ },
53
+ {
54
+ "epoch": 0.2000128460402081,
55
+ "eval_loss": 0.164720818400383,
56
+ "eval_runtime": 7.2965,
57
+ "eval_samples_per_second": 68.526,
58
+ "eval_steps_per_second": 8.634,
59
+ "step": 3114
60
+ },
61
+ {
62
+ "epoch": 0.2248057036418524,
63
+ "grad_norm": 0.6251304745674133,
64
+ "learning_rate": 0.00039251289956537567,
65
+ "loss": 0.1745,
66
+ "step": 3500
67
+ },
68
+ {
69
+ "epoch": 0.25692080416211704,
70
+ "grad_norm": 0.46483030915260315,
71
+ "learning_rate": 0.0003914445372214015,
72
+ "loss": 0.1701,
73
+ "step": 4000
74
+ },
75
+ {
76
+ "epoch": 0.28903590468238166,
77
+ "grad_norm": 0.477145254611969,
78
+ "learning_rate": 0.000390374033870726,
79
+ "loss": 0.1692,
80
+ "step": 4500
81
+ },
82
+ {
83
+ "epoch": 0.3211510052026463,
84
+ "grad_norm": 0.4863433539867401,
85
+ "learning_rate": 0.00038930353052005055,
86
+ "loss": 0.1633,
87
+ "step": 5000
88
+ },
89
+ {
90
+ "epoch": 0.3532661057229109,
91
+ "grad_norm": 0.3692472279071808,
92
+ "learning_rate": 0.000388233027169375,
93
+ "loss": 0.1623,
94
+ "step": 5500
95
+ },
96
+ {
97
+ "epoch": 0.38538120624317557,
98
+ "grad_norm": 0.5566384792327881,
99
+ "learning_rate": 0.00038716252381869955,
100
+ "loss": 0.1601,
101
+ "step": 6000
102
+ },
103
+ {
104
+ "epoch": 0.4000256920804162,
105
+ "eval_loss": 0.14738531410694122,
106
+ "eval_runtime": 7.3011,
107
+ "eval_samples_per_second": 68.483,
108
+ "eval_steps_per_second": 8.629,
109
+ "step": 6228
110
+ },
111
+ {
112
+ "epoch": 0.4174963067634402,
113
+ "grad_norm": 0.5637210011482239,
114
+ "learning_rate": 0.0003860920204680241,
115
+ "loss": 0.1569,
116
+ "step": 6500
117
+ },
118
+ {
119
+ "epoch": 0.4496114072837048,
120
+ "grad_norm": 0.34262317419052124,
121
+ "learning_rate": 0.00038502365812404996,
122
+ "loss": 0.1549,
123
+ "step": 7000
124
+ },
125
+ {
126
+ "epoch": 0.4817265078039694,
127
+ "grad_norm": 0.8737928867340088,
128
+ "learning_rate": 0.00038395529578007584,
129
+ "loss": 0.1558,
130
+ "step": 7500
131
+ },
132
+ {
133
+ "epoch": 0.5138416083242341,
134
+ "grad_norm": 0.3039982318878174,
135
+ "learning_rate": 0.00038288479242940037,
136
+ "loss": 0.1519,
137
+ "step": 8000
138
+ },
139
+ {
140
+ "epoch": 0.5459567088444987,
141
+ "grad_norm": 0.4347224831581116,
142
+ "learning_rate": 0.00038181428907872484,
143
+ "loss": 0.1505,
144
+ "step": 8500
145
+ },
146
+ {
147
+ "epoch": 0.5780718093647633,
148
+ "grad_norm": 0.39854001998901367,
149
+ "learning_rate": 0.00038074378572804937,
150
+ "loss": 0.1496,
151
+ "step": 9000
152
+ },
153
+ {
154
+ "epoch": 0.6000385381206244,
155
+ "eval_loss": 0.13884712755680084,
156
+ "eval_runtime": 7.294,
157
+ "eval_samples_per_second": 68.55,
158
+ "eval_steps_per_second": 8.637,
159
+ "step": 9342
160
+ },
161
+ {
162
+ "epoch": 0.610186909885028,
163
+ "grad_norm": 0.6563107967376709,
164
+ "learning_rate": 0.0003796732823773739,
165
+ "loss": 0.1472,
166
+ "step": 9500
167
+ },
168
+ {
169
+ "epoch": 0.6423020104052926,
170
+ "grad_norm": 0.3582533597946167,
171
+ "learning_rate": 0.00037860277902669837,
172
+ "loss": 0.1457,
173
+ "step": 10000
174
+ },
175
+ {
176
+ "epoch": 0.6744171109255572,
177
+ "grad_norm": 0.6487281322479248,
178
+ "learning_rate": 0.0003775322756760229,
179
+ "loss": 0.1484,
180
+ "step": 10500
181
+ },
182
+ {
183
+ "epoch": 0.7065322114458218,
184
+ "grad_norm": 0.6346992254257202,
185
+ "learning_rate": 0.0003764617723253474,
186
+ "loss": 0.1453,
187
+ "step": 11000
188
+ },
189
+ {
190
+ "epoch": 0.7386473119660865,
191
+ "grad_norm": 0.5041869878768921,
192
+ "learning_rate": 0.0003753912689746719,
193
+ "loss": 0.1454,
194
+ "step": 11500
195
+ },
196
+ {
197
+ "epoch": 0.7707624124863511,
198
+ "grad_norm": 0.3400748372077942,
199
+ "learning_rate": 0.0003743207656239964,
200
+ "loss": 0.1441,
201
+ "step": 12000
202
+ },
203
+ {
204
+ "epoch": 0.8000513841608324,
205
+ "eval_loss": 0.13148044049739838,
206
+ "eval_runtime": 7.2929,
207
+ "eval_samples_per_second": 68.56,
208
+ "eval_steps_per_second": 8.639,
209
+ "step": 12456
210
+ },
211
+ {
212
+ "epoch": 0.8028775130066157,
213
+ "grad_norm": 0.32738161087036133,
214
+ "learning_rate": 0.00037325026227332095,
215
+ "loss": 0.1445,
216
+ "step": 12500
217
+ },
218
+ {
219
+ "epoch": 0.8349926135268804,
220
+ "grad_norm": 0.3660726845264435,
221
+ "learning_rate": 0.0003721818999293468,
222
+ "loss": 0.1424,
223
+ "step": 13000
224
+ },
225
+ {
226
+ "epoch": 0.8671077140471449,
227
+ "grad_norm": 0.5750604867935181,
228
+ "learning_rate": 0.000371115678592074,
229
+ "loss": 0.1469,
230
+ "step": 13500
231
+ },
232
+ {
233
+ "epoch": 0.8992228145674096,
234
+ "grad_norm": 0.35107406973838806,
235
+ "learning_rate": 0.00037004517524139854,
236
+ "loss": 0.147,
237
+ "step": 14000
238
+ },
239
+ {
240
+ "epoch": 0.9313379150876743,
241
+ "grad_norm": 0.4457557499408722,
242
+ "learning_rate": 0.000368974671890723,
243
+ "loss": 0.1416,
244
+ "step": 14500
245
+ },
246
+ {
247
+ "epoch": 0.9634530156079388,
248
+ "grad_norm": 0.4570375382900238,
249
+ "learning_rate": 0.00036790416854004754,
250
+ "loss": 0.1391,
251
+ "step": 15000
252
+ },
253
+ {
254
+ "epoch": 0.9955681161282035,
255
+ "grad_norm": 0.5502489805221558,
256
+ "learning_rate": 0.00036683366518937207,
257
+ "loss": 0.1408,
258
+ "step": 15500
259
+ },
260
+ {
261
+ "epoch": 1.0000642302010405,
262
+ "eval_loss": 0.1298205852508545,
263
+ "eval_runtime": 7.3714,
264
+ "eval_samples_per_second": 67.829,
265
+ "eval_steps_per_second": 8.547,
266
+ "step": 15570
267
+ },
268
+ {
269
+ "epoch": 1.0276832166484682,
270
+ "grad_norm": 0.41828590631484985,
271
+ "learning_rate": 0.0003657631618386966,
272
+ "loss": 0.1312,
273
+ "step": 16000
274
+ },
275
+ {
276
+ "epoch": 1.0597983171687329,
277
+ "grad_norm": 0.4035309851169586,
278
+ "learning_rate": 0.00036469265848802107,
279
+ "loss": 0.1298,
280
+ "step": 16500
281
+ },
282
+ {
283
+ "epoch": 1.0919134176889973,
284
+ "grad_norm": 0.38738253712654114,
285
+ "learning_rate": 0.0003636221551373456,
286
+ "loss": 0.1289,
287
+ "step": 17000
288
+ },
289
+ {
290
+ "epoch": 1.124028518209262,
291
+ "grad_norm": 0.4483577609062195,
292
+ "learning_rate": 0.0003625516517866701,
293
+ "loss": 0.1308,
294
+ "step": 17500
295
+ },
296
+ {
297
+ "epoch": 1.1561436187295266,
298
+ "grad_norm": 0.36913159489631653,
299
+ "learning_rate": 0.0003614811484359946,
300
+ "loss": 0.1313,
301
+ "step": 18000
302
+ },
303
+ {
304
+ "epoch": 1.1882587192497913,
305
+ "grad_norm": 0.4135682284832001,
306
+ "learning_rate": 0.0003604106450853191,
307
+ "loss": 0.1308,
308
+ "step": 18500
309
+ },
310
+ {
311
+ "epoch": 1.2000770762412487,
312
+ "eval_loss": 0.12298451364040375,
313
+ "eval_runtime": 7.2977,
314
+ "eval_samples_per_second": 68.515,
315
+ "eval_steps_per_second": 8.633,
316
+ "step": 18684
317
+ },
318
+ {
319
+ "epoch": 1.2203738197700558,
320
+ "grad_norm": 0.5066474676132202,
321
+ "learning_rate": 0.00035934014173464365,
322
+ "loss": 0.1294,
323
+ "step": 19000
324
+ },
325
+ {
326
+ "epoch": 1.2524889202903204,
327
+ "grad_norm": 0.35160061717033386,
328
+ "learning_rate": 0.0003582696383839681,
329
+ "loss": 0.1297,
330
+ "step": 19500
331
+ },
332
+ {
333
+ "epoch": 1.2846040208105851,
334
+ "grad_norm": 0.5103857517242432,
335
+ "learning_rate": 0.00035719913503329265,
336
+ "loss": 0.1307,
337
+ "step": 20000
338
+ },
339
+ {
340
+ "epoch": 1.3167191213308498,
341
+ "grad_norm": 0.35336533188819885,
342
+ "learning_rate": 0.0003561286316826172,
343
+ "loss": 0.1275,
344
+ "step": 20500
345
+ },
346
+ {
347
+ "epoch": 1.3488342218511145,
348
+ "grad_norm": 0.3233303725719452,
349
+ "learning_rate": 0.0003550581283319417,
350
+ "loss": 0.1247,
351
+ "step": 21000
352
+ },
353
+ {
354
+ "epoch": 1.3809493223713791,
355
+ "grad_norm": 0.4102284014225006,
356
+ "learning_rate": 0.0003539897659879676,
357
+ "loss": 0.128,
358
+ "step": 21500
359
+ },
360
+ {
361
+ "epoch": 1.4000899222814567,
362
+ "eval_loss": 0.12551982700824738,
363
+ "eval_runtime": 7.3096,
364
+ "eval_samples_per_second": 68.403,
365
+ "eval_steps_per_second": 8.619,
366
+ "step": 21798
367
+ },
368
+ {
369
+ "epoch": 1.4130644228916436,
370
+ "grad_norm": 0.43554574251174927,
371
+ "learning_rate": 0.0003529192626372921,
372
+ "loss": 0.1268,
373
+ "step": 22000
374
+ },
375
+ {
376
+ "epoch": 1.4451795234119083,
377
+ "grad_norm": 0.4067144989967346,
378
+ "learning_rate": 0.0003518487592866166,
379
+ "loss": 0.1265,
380
+ "step": 22500
381
+ },
382
+ {
383
+ "epoch": 1.477294623932173,
384
+ "grad_norm": 0.5447452664375305,
385
+ "learning_rate": 0.0003507782559359411,
386
+ "loss": 0.1261,
387
+ "step": 23000
388
+ },
389
+ {
390
+ "epoch": 1.5094097244524374,
391
+ "grad_norm": 0.5342037677764893,
392
+ "learning_rate": 0.000349709893591967,
393
+ "loss": 0.1249,
394
+ "step": 23500
395
+ },
396
+ {
397
+ "epoch": 1.541524824972702,
398
+ "grad_norm": 0.5300814509391785,
399
+ "learning_rate": 0.00034863939024129146,
400
+ "loss": 0.1255,
401
+ "step": 24000
402
+ },
403
+ {
404
+ "epoch": 1.5736399254929667,
405
+ "grad_norm": 0.4534071981906891,
406
+ "learning_rate": 0.00034757102789731735,
407
+ "loss": 0.1281,
408
+ "step": 24500
409
+ },
410
+ {
411
+ "epoch": 1.6001027683216649,
412
+ "eval_loss": 0.11974115669727325,
413
+ "eval_runtime": 7.3112,
414
+ "eval_samples_per_second": 68.388,
415
+ "eval_steps_per_second": 8.617,
416
+ "step": 24912
417
+ },
418
+ {
419
+ "epoch": 1.6057550260132314,
420
+ "grad_norm": 0.39113229513168335,
421
+ "learning_rate": 0.0003465005245466419,
422
+ "loss": 0.1269,
423
+ "step": 25000
424
+ },
425
+ {
426
+ "epoch": 1.637870126533496,
427
+ "grad_norm": 0.5367024540901184,
428
+ "learning_rate": 0.00034543002119596635,
429
+ "loss": 0.1268,
430
+ "step": 25500
431
+ },
432
+ {
433
+ "epoch": 1.6699852270537607,
434
+ "grad_norm": 0.3257845640182495,
435
+ "learning_rate": 0.0003443595178452909,
436
+ "loss": 0.1236,
437
+ "step": 26000
438
+ },
439
+ {
440
+ "epoch": 1.7021003275740254,
441
+ "grad_norm": 0.4521988034248352,
442
+ "learning_rate": 0.0003432890144946154,
443
+ "loss": 0.1234,
444
+ "step": 26500
445
+ },
446
+ {
447
+ "epoch": 1.73421542809429,
448
+ "grad_norm": 0.4063723683357239,
449
+ "learning_rate": 0.00034222065215064123,
450
+ "loss": 0.1238,
451
+ "step": 27000
452
+ },
453
+ {
454
+ "epoch": 1.7663305286145545,
455
+ "grad_norm": 0.41461870074272156,
456
+ "learning_rate": 0.00034115014879996576,
457
+ "loss": 0.123,
458
+ "step": 27500
459
+ },
460
+ {
461
+ "epoch": 1.7984456291348192,
462
+ "grad_norm": 0.3680518567562103,
463
+ "learning_rate": 0.0003400796454492903,
464
+ "loss": 0.1239,
465
+ "step": 28000
466
+ },
467
+ {
468
+ "epoch": 1.8001156143618728,
469
+ "eval_loss": 0.11659311503171921,
470
+ "eval_runtime": 7.3186,
471
+ "eval_samples_per_second": 68.319,
472
+ "eval_steps_per_second": 8.608,
473
+ "step": 28026
474
+ },
475
+ {
476
+ "epoch": 1.8305607296550839,
477
+ "grad_norm": 0.4289158284664154,
478
+ "learning_rate": 0.0003390091420986148,
479
+ "loss": 0.1265,
480
+ "step": 28500
481
+ },
482
+ {
483
+ "epoch": 1.8626758301753483,
484
+ "grad_norm": 0.6132258772850037,
485
+ "learning_rate": 0.0003379386387479393,
486
+ "loss": 0.1231,
487
+ "step": 29000
488
+ },
489
+ {
490
+ "epoch": 1.894790930695613,
491
+ "grad_norm": 0.6142964959144592,
492
+ "learning_rate": 0.0003368681353972638,
493
+ "loss": 0.1242,
494
+ "step": 29500
495
+ },
496
+ {
497
+ "epoch": 1.9269060312158777,
498
+ "grad_norm": 0.29933828115463257,
499
+ "learning_rate": 0.00033579763204658834,
500
+ "loss": 0.1224,
501
+ "step": 30000
502
+ },
503
+ {
504
+ "epoch": 1.9590211317361423,
505
+ "grad_norm": 0.4025266170501709,
506
+ "learning_rate": 0.0003347271286959128,
507
+ "loss": 0.123,
508
+ "step": 30500
509
+ },
510
+ {
511
+ "epoch": 1.991136232256407,
512
+ "grad_norm": 0.4255163073539734,
513
+ "learning_rate": 0.00033365662534523734,
514
+ "loss": 0.1226,
515
+ "step": 31000
516
+ },
517
+ {
518
+ "epoch": 2.000128460402081,
519
+ "eval_loss": 0.11819276958703995,
520
+ "eval_runtime": 7.4375,
521
+ "eval_samples_per_second": 67.227,
522
+ "eval_steps_per_second": 8.471,
523
+ "step": 31140
524
+ },
525
+ {
526
+ "epoch": 2.0232513327766717,
527
+ "grad_norm": 0.5175437927246094,
528
+ "learning_rate": 0.00033258612199456186,
529
+ "loss": 0.1163,
530
+ "step": 31500
531
+ },
532
+ {
533
+ "epoch": 2.0553664332969364,
534
+ "grad_norm": 0.43255355954170227,
535
+ "learning_rate": 0.00033151561864388634,
536
+ "loss": 0.1146,
537
+ "step": 32000
538
+ },
539
+ {
540
+ "epoch": 2.087481533817201,
541
+ "grad_norm": 0.506144642829895,
542
+ "learning_rate": 0.00033044511529321086,
543
+ "loss": 0.1129,
544
+ "step": 32500
545
+ },
546
+ {
547
+ "epoch": 2.1195966343374657,
548
+ "grad_norm": 0.5078917145729065,
549
+ "learning_rate": 0.0003293746119425354,
550
+ "loss": 0.1145,
551
+ "step": 33000
552
+ },
553
+ {
554
+ "epoch": 2.15171173485773,
555
+ "grad_norm": 0.507586658000946,
556
+ "learning_rate": 0.0003283041085918599,
557
+ "loss": 0.1123,
558
+ "step": 33500
559
+ },
560
+ {
561
+ "epoch": 2.1838268353779946,
562
+ "grad_norm": 0.4673985242843628,
563
+ "learning_rate": 0.00032723574624788574,
564
+ "loss": 0.1136,
565
+ "step": 34000
566
+ },
567
+ {
568
+ "epoch": 2.2001413064422892,
569
+ "eval_loss": 0.11647045612335205,
570
+ "eval_runtime": 7.3242,
571
+ "eval_samples_per_second": 68.267,
572
+ "eval_steps_per_second": 8.602,
573
+ "step": 34254
574
+ },
575
+ {
576
+ "epoch": 2.2159419358982593,
577
+ "grad_norm": 0.5816234350204468,
578
+ "learning_rate": 0.0003261673839039116,
579
+ "loss": 0.1162,
580
+ "step": 34500
581
+ },
582
+ {
583
+ "epoch": 2.248057036418524,
584
+ "grad_norm": 0.3159160912036896,
585
+ "learning_rate": 0.00032509688055323615,
586
+ "loss": 0.1177,
587
+ "step": 35000
588
+ },
589
+ {
590
+ "epoch": 2.2801721369387886,
591
+ "grad_norm": 0.5306550860404968,
592
+ "learning_rate": 0.0003240263772025606,
593
+ "loss": 0.1144,
594
+ "step": 35500
595
+ },
596
+ {
597
+ "epoch": 2.3122872374590533,
598
+ "grad_norm": 0.5449818968772888,
599
+ "learning_rate": 0.0003229558738518852,
600
+ "loss": 0.113,
601
+ "step": 36000
602
+ },
603
+ {
604
+ "epoch": 2.344402337979318,
605
+ "grad_norm": 0.41796988248825073,
606
+ "learning_rate": 0.0003218853705012097,
607
+ "loss": 0.114,
608
+ "step": 36500
609
+ },
610
+ {
611
+ "epoch": 2.3765174384995826,
612
+ "grad_norm": 0.5479010343551636,
613
+ "learning_rate": 0.00032081700815723556,
614
+ "loss": 0.1148,
615
+ "step": 37000
616
+ },
617
+ {
618
+ "epoch": 2.4001541524824974,
619
+ "eval_loss": 0.11464451998472214,
620
+ "eval_runtime": 7.324,
621
+ "eval_samples_per_second": 68.269,
622
+ "eval_steps_per_second": 8.602,
623
+ "step": 37368
624
+ },
625
+ {
626
+ "epoch": 2.4086325390198473,
627
+ "grad_norm": 0.34335535764694214,
628
+ "learning_rate": 0.0003197465048065601,
629
+ "loss": 0.1144,
630
+ "step": 37500
631
+ },
632
+ {
633
+ "epoch": 2.4407476395401115,
634
+ "grad_norm": 0.4916399419307709,
635
+ "learning_rate": 0.00031867600145588456,
636
+ "loss": 0.1134,
637
+ "step": 38000
638
+ },
639
+ {
640
+ "epoch": 2.472862740060376,
641
+ "grad_norm": 0.5686454772949219,
642
+ "learning_rate": 0.0003176054981052091,
643
+ "loss": 0.1116,
644
+ "step": 38500
645
+ },
646
+ {
647
+ "epoch": 2.504977840580641,
648
+ "grad_norm": 0.3755095601081848,
649
+ "learning_rate": 0.0003165349947545336,
650
+ "loss": 0.1128,
651
+ "step": 39000
652
+ },
653
+ {
654
+ "epoch": 2.5370929411009056,
655
+ "grad_norm": 0.4691629409790039,
656
+ "learning_rate": 0.00031546449140385814,
657
+ "loss": 0.1123,
658
+ "step": 39500
659
+ },
660
+ {
661
+ "epoch": 2.5692080416211702,
662
+ "grad_norm": 0.5287884473800659,
663
+ "learning_rate": 0.0003143939880531826,
664
+ "loss": 0.1154,
665
+ "step": 40000
666
+ },
667
+ {
668
+ "epoch": 2.600166998522705,
669
+ "eval_loss": 0.10910294204950333,
670
+ "eval_runtime": 7.3339,
671
+ "eval_samples_per_second": 68.176,
672
+ "eval_steps_per_second": 8.59,
673
+ "step": 40482
674
+ },
675
+ {
676
+ "epoch": 2.601323142141435,
677
+ "grad_norm": 0.5030823945999146,
678
+ "learning_rate": 0.00031332348470250714,
679
+ "loss": 0.1145,
680
+ "step": 40500
681
+ },
682
+ {
683
+ "epoch": 2.6334382426616996,
684
+ "grad_norm": 0.6581201553344727,
685
+ "learning_rate": 0.00031225298135183167,
686
+ "loss": 0.1127,
687
+ "step": 41000
688
+ },
689
+ {
690
+ "epoch": 2.6655533431819642,
691
+ "grad_norm": 0.5677359700202942,
692
+ "learning_rate": 0.0003111846190078575,
693
+ "loss": 0.1121,
694
+ "step": 41500
695
+ },
696
+ {
697
+ "epoch": 2.697668443702229,
698
+ "grad_norm": 0.44164523482322693,
699
+ "learning_rate": 0.0003101162566638834,
700
+ "loss": 0.1106,
701
+ "step": 42000
702
+ },
703
+ {
704
+ "epoch": 2.729783544222493,
705
+ "grad_norm": 0.5856674313545227,
706
+ "learning_rate": 0.0003090457533132079,
707
+ "loss": 0.1102,
708
+ "step": 42500
709
+ },
710
+ {
711
+ "epoch": 2.7618986447427583,
712
+ "grad_norm": 0.41598838567733765,
713
+ "learning_rate": 0.0003079752499625324,
714
+ "loss": 0.1123,
715
+ "step": 43000
716
+ },
717
+ {
718
+ "epoch": 2.7940137452630225,
719
+ "grad_norm": 0.5005472898483276,
720
+ "learning_rate": 0.0003069047466118569,
721
+ "loss": 0.1142,
722
+ "step": 43500
723
+ },
724
+ {
725
+ "epoch": 2.8001798445629134,
726
+ "eval_loss": 0.10679365694522858,
727
+ "eval_runtime": 7.3241,
728
+ "eval_samples_per_second": 68.267,
729
+ "eval_steps_per_second": 8.602,
730
+ "step": 43596
731
+ },
732
+ {
733
+ "epoch": 2.826128845783287,
734
+ "grad_norm": 0.615060031414032,
735
+ "learning_rate": 0.00030583424326118143,
736
+ "loss": 0.1125,
737
+ "step": 44000
738
+ },
739
+ {
740
+ "epoch": 2.858243946303552,
741
+ "grad_norm": 0.5374606251716614,
742
+ "learning_rate": 0.0003047637399105059,
743
+ "loss": 0.1121,
744
+ "step": 44500
745
+ },
746
+ {
747
+ "epoch": 2.8903590468238165,
748
+ "grad_norm": 0.44819122552871704,
749
+ "learning_rate": 0.00030369323655983043,
750
+ "loss": 0.1127,
751
+ "step": 45000
752
+ },
753
+ {
754
+ "epoch": 2.922474147344081,
755
+ "grad_norm": 0.4970841705799103,
756
+ "learning_rate": 0.00030262273320915496,
757
+ "loss": 0.1104,
758
+ "step": 45500
759
+ },
760
+ {
761
+ "epoch": 2.954589247864346,
762
+ "grad_norm": 0.4294310510158539,
763
+ "learning_rate": 0.0003015543708651808,
764
+ "loss": 0.1108,
765
+ "step": 46000
766
+ },
767
+ {
768
+ "epoch": 2.9867043483846105,
769
+ "grad_norm": 0.8992510437965393,
770
+ "learning_rate": 0.0003004838675145053,
771
+ "loss": 0.1092,
772
+ "step": 46500
773
+ },
774
+ {
775
+ "epoch": 3.0001926906031215,
776
+ "eval_loss": 0.1059611514210701,
777
+ "eval_runtime": 7.487,
778
+ "eval_samples_per_second": 66.782,
779
+ "eval_steps_per_second": 8.415,
780
+ "step": 46710
781
+ },
782
+ {
783
+ "epoch": 3.018819448904875,
784
+ "grad_norm": 0.44859662652015686,
785
+ "learning_rate": 0.00029941336416382984,
786
+ "loss": 0.1058,
787
+ "step": 47000
788
+ },
789
+ {
790
+ "epoch": 3.05093454942514,
791
+ "grad_norm": 0.47559085488319397,
792
+ "learning_rate": 0.00029834286081315437,
793
+ "loss": 0.1032,
794
+ "step": 47500
795
+ },
796
+ {
797
+ "epoch": 3.0830496499454045,
798
+ "grad_norm": 0.47841718792915344,
799
+ "learning_rate": 0.00029727235746247884,
800
+ "loss": 0.101,
801
+ "step": 48000
802
+ },
803
+ {
804
+ "epoch": 3.1151647504656688,
805
+ "grad_norm": 0.40354153513908386,
806
+ "learning_rate": 0.0002962039951185047,
807
+ "loss": 0.1011,
808
+ "step": 48500
809
+ },
810
+ {
811
+ "epoch": 3.1472798509859334,
812
+ "grad_norm": 0.4232326149940491,
813
+ "learning_rate": 0.00029513563277453055,
814
+ "loss": 0.1022,
815
+ "step": 49000
816
+ },
817
+ {
818
+ "epoch": 3.179394951506198,
819
+ "grad_norm": 0.3321347236633301,
820
+ "learning_rate": 0.0002940651294238551,
821
+ "loss": 0.1035,
822
+ "step": 49500
823
+ },
824
+ {
825
+ "epoch": 3.2002055366433297,
826
+ "eval_loss": 0.10808317363262177,
827
+ "eval_runtime": 7.3206,
828
+ "eval_samples_per_second": 68.301,
829
+ "eval_steps_per_second": 8.606,
830
+ "step": 49824
831
+ },
832
+ {
833
+ "epoch": 3.211510052026463,
834
+ "grad_norm": 0.3930283188819885,
835
+ "learning_rate": 0.0002929946260731796,
836
+ "loss": 0.1031,
837
+ "step": 50000
838
+ },
839
+ {
840
+ "epoch": 3.2436251525467275,
841
+ "grad_norm": 0.3710777759552002,
842
+ "learning_rate": 0.00029192412272250413,
843
+ "loss": 0.1037,
844
+ "step": 50500
845
+ },
846
+ {
847
+ "epoch": 3.275740253066992,
848
+ "grad_norm": 0.4590394198894501,
849
+ "learning_rate": 0.00029085361937182866,
850
+ "loss": 0.1014,
851
+ "step": 51000
852
+ },
853
+ {
854
+ "epoch": 3.307855353587257,
855
+ "grad_norm": 0.5470544099807739,
856
+ "learning_rate": 0.0002897831160211532,
857
+ "loss": 0.1044,
858
+ "step": 51500
859
+ },
860
+ {
861
+ "epoch": 3.3399704541075215,
862
+ "grad_norm": 0.7178152203559875,
863
+ "learning_rate": 0.0002887126126704777,
864
+ "loss": 0.1017,
865
+ "step": 52000
866
+ },
867
+ {
868
+ "epoch": 3.372085554627786,
869
+ "grad_norm": 0.5023342370986938,
870
+ "learning_rate": 0.0002876421093198022,
871
+ "loss": 0.1045,
872
+ "step": 52500
873
+ },
874
+ {
875
+ "epoch": 3.400218382683538,
876
+ "eval_loss": 0.1036703959107399,
877
+ "eval_runtime": 7.3368,
878
+ "eval_samples_per_second": 68.15,
879
+ "eval_steps_per_second": 8.587,
880
+ "step": 52938
881
+ },
882
+ {
883
+ "epoch": 3.404200655148051,
884
+ "grad_norm": 0.49040645360946655,
885
+ "learning_rate": 0.0002865716059691267,
886
+ "loss": 0.1057,
887
+ "step": 53000
888
+ },
889
+ {
890
+ "epoch": 3.436315755668315,
891
+ "grad_norm": 0.4539451599121094,
892
+ "learning_rate": 0.0002855053846318539,
893
+ "loss": 0.1045,
894
+ "step": 53500
895
+ },
896
+ {
897
+ "epoch": 3.4684308561885797,
898
+ "grad_norm": 0.3792193830013275,
899
+ "learning_rate": 0.0002844348812811784,
900
+ "loss": 0.1051,
901
+ "step": 54000
902
+ },
903
+ {
904
+ "epoch": 3.5005459567088444,
905
+ "grad_norm": 0.5674263834953308,
906
+ "learning_rate": 0.00028336437793050295,
907
+ "loss": 0.1052,
908
+ "step": 54500
909
+ },
910
+ {
911
+ "epoch": 3.532661057229109,
912
+ "grad_norm": 0.550486147403717,
913
+ "learning_rate": 0.0002822938745798275,
914
+ "loss": 0.1044,
915
+ "step": 55000
916
+ },
917
+ {
918
+ "epoch": 3.5647761577493737,
919
+ "grad_norm": 0.42234089970588684,
920
+ "learning_rate": 0.00028122337122915195,
921
+ "loss": 0.1043,
922
+ "step": 55500
923
+ },
924
+ {
925
+ "epoch": 3.5968912582696384,
926
+ "grad_norm": 0.5359342098236084,
927
+ "learning_rate": 0.0002801528678784765,
928
+ "loss": 0.1027,
929
+ "step": 56000
930
+ },
931
+ {
932
+ "epoch": 3.6002312287237457,
933
+ "eval_loss": 0.10499866306781769,
934
+ "eval_runtime": 7.3501,
935
+ "eval_samples_per_second": 68.026,
936
+ "eval_steps_per_second": 8.571,
937
+ "step": 56052
938
+ },
939
+ {
940
+ "epoch": 3.629006358789903,
941
+ "grad_norm": 0.5378216505050659,
942
+ "learning_rate": 0.000279082364527801,
943
+ "loss": 0.1037,
944
+ "step": 56500
945
+ },
946
+ {
947
+ "epoch": 3.6611214593101677,
948
+ "grad_norm": 0.47160041332244873,
949
+ "learning_rate": 0.0002780118611771255,
950
+ "loss": 0.1032,
951
+ "step": 57000
952
+ },
953
+ {
954
+ "epoch": 3.6932365598304324,
955
+ "grad_norm": 0.49247318506240845,
956
+ "learning_rate": 0.00027694135782645,
957
+ "loss": 0.1005,
958
+ "step": 57500
959
+ },
960
+ {
961
+ "epoch": 3.7253516603506966,
962
+ "grad_norm": 0.46599718928337097,
963
+ "learning_rate": 0.00027587085447577453,
964
+ "loss": 0.1037,
965
+ "step": 58000
966
+ },
967
+ {
968
+ "epoch": 3.7574667608709618,
969
+ "grad_norm": 0.44333142042160034,
970
+ "learning_rate": 0.00027480249213180036,
971
+ "loss": 0.0991,
972
+ "step": 58500
973
+ },
974
+ {
975
+ "epoch": 3.789581861391226,
976
+ "grad_norm": 0.42183053493499756,
977
+ "learning_rate": 0.0002737319887811249,
978
+ "loss": 0.1025,
979
+ "step": 59000
980
+ },
981
+ {
982
+ "epoch": 3.800244074763954,
983
+ "eval_loss": 0.10221809148788452,
984
+ "eval_runtime": 7.3337,
985
+ "eval_samples_per_second": 68.178,
986
+ "eval_steps_per_second": 8.59,
987
+ "step": 59166
988
+ },
989
+ {
990
+ "epoch": 3.8216969619114907,
991
+ "grad_norm": 0.4511483907699585,
992
+ "learning_rate": 0.0002726614854304494,
993
+ "loss": 0.1046,
994
+ "step": 59500
995
+ },
996
+ {
997
+ "epoch": 3.8538120624317553,
998
+ "grad_norm": 0.6245924830436707,
999
+ "learning_rate": 0.0002715909820797739,
1000
+ "loss": 0.0996,
1001
+ "step": 60000
1002
+ },
1003
+ {
1004
+ "epoch": 3.88592716295202,
1005
+ "grad_norm": 0.43624454736709595,
1006
+ "learning_rate": 0.00027052261973579977,
1007
+ "loss": 0.1034,
1008
+ "step": 60500
1009
+ },
1010
+ {
1011
+ "epoch": 3.9180422634722847,
1012
+ "grad_norm": 0.47011294960975647,
1013
+ "learning_rate": 0.0002694521163851243,
1014
+ "loss": 0.1039,
1015
+ "step": 61000
1016
+ },
1017
+ {
1018
+ "epoch": 3.9501573639925494,
1019
+ "grad_norm": 0.4194742441177368,
1020
+ "learning_rate": 0.0002683816130344488,
1021
+ "loss": 0.1048,
1022
+ "step": 61500
1023
+ },
1024
+ {
1025
+ "epoch": 3.982272464512814,
1026
+ "grad_norm": 0.5618862509727478,
1027
+ "learning_rate": 0.0002673111096837733,
1028
+ "loss": 0.102,
1029
+ "step": 62000
1030
+ },
1031
+ {
1032
+ "epoch": 4.000256920804162,
1033
+ "eval_loss": 0.10089205950498581,
1034
+ "eval_runtime": 7.4502,
1035
+ "eval_samples_per_second": 67.113,
1036
+ "eval_steps_per_second": 8.456,
1037
+ "step": 62280
1038
+ },
1039
+ {
1040
+ "epoch": 4.014387565033078,
1041
+ "grad_norm": 0.60516756772995,
1042
+ "learning_rate": 0.0002662406063330978,
1043
+ "loss": 0.0979,
1044
+ "step": 62500
1045
+ },
1046
+ {
1047
+ "epoch": 4.046502665553343,
1048
+ "grad_norm": 0.5307052731513977,
1049
+ "learning_rate": 0.00026517010298242235,
1050
+ "loss": 0.0938,
1051
+ "step": 63000
1052
+ },
1053
+ {
1054
+ "epoch": 4.078617766073608,
1055
+ "grad_norm": 0.513951301574707,
1056
+ "learning_rate": 0.0002641017406384482,
1057
+ "loss": 0.0948,
1058
+ "step": 63500
1059
+ },
1060
+ {
1061
+ "epoch": 4.110732866593873,
1062
+ "grad_norm": 0.4734605848789215,
1063
+ "learning_rate": 0.0002630312372877727,
1064
+ "loss": 0.0956,
1065
+ "step": 64000
1066
+ },
1067
+ {
1068
+ "epoch": 4.142847967114137,
1069
+ "grad_norm": 0.3525328040122986,
1070
+ "learning_rate": 0.00026196073393709723,
1071
+ "loss": 0.0935,
1072
+ "step": 64500
1073
+ },
1074
+ {
1075
+ "epoch": 4.174963067634402,
1076
+ "grad_norm": 0.6330441832542419,
1077
+ "learning_rate": 0.00026089023058642176,
1078
+ "loss": 0.0943,
1079
+ "step": 65000
1080
+ },
1081
+ {
1082
+ "epoch": 4.20026976684437,
1083
+ "eval_loss": 0.09817148745059967,
1084
+ "eval_runtime": 7.3411,
1085
+ "eval_samples_per_second": 68.11,
1086
+ "eval_steps_per_second": 8.582,
1087
+ "step": 65394
1088
+ },
1089
+ {
1090
+ "epoch": 4.207078168154666,
1091
+ "grad_norm": 0.3115638494491577,
1092
+ "learning_rate": 0.0002598197272357463,
1093
+ "loss": 0.095,
1094
+ "step": 65500
1095
+ },
1096
+ {
1097
+ "epoch": 4.239193268674931,
1098
+ "grad_norm": 0.3824947774410248,
1099
+ "learning_rate": 0.0002587492238850708,
1100
+ "loss": 0.0943,
1101
+ "step": 66000
1102
+ },
1103
+ {
1104
+ "epoch": 4.271308369195196,
1105
+ "grad_norm": 0.547127366065979,
1106
+ "learning_rate": 0.00025768086154109664,
1107
+ "loss": 0.0956,
1108
+ "step": 66500
1109
+ },
1110
+ {
1111
+ "epoch": 4.30342346971546,
1112
+ "grad_norm": 0.39464259147644043,
1113
+ "learning_rate": 0.00025661035819042117,
1114
+ "loss": 0.0951,
1115
+ "step": 67000
1116
+ },
1117
+ {
1118
+ "epoch": 4.335538570235725,
1119
+ "grad_norm": 0.465304434299469,
1120
+ "learning_rate": 0.0002555398548397457,
1121
+ "loss": 0.0935,
1122
+ "step": 67500
1123
+ },
1124
+ {
1125
+ "epoch": 4.367653670755989,
1126
+ "grad_norm": 0.5602743625640869,
1127
+ "learning_rate": 0.00025446935148907017,
1128
+ "loss": 0.0972,
1129
+ "step": 68000
1130
+ },
1131
+ {
1132
+ "epoch": 4.399768771276254,
1133
+ "grad_norm": 0.3761026859283447,
1134
+ "learning_rate": 0.00025340098914509605,
1135
+ "loss": 0.0948,
1136
+ "step": 68500
1137
+ },
1138
+ {
1139
+ "epoch": 4.4002826128845784,
1140
+ "eval_loss": 0.09914599359035492,
1141
+ "eval_runtime": 7.3361,
1142
+ "eval_samples_per_second": 68.156,
1143
+ "eval_steps_per_second": 8.588,
1144
+ "step": 68508
1145
+ },
1146
+ {
1147
+ "epoch": 4.4318838717965185,
1148
+ "grad_norm": 0.5205414891242981,
1149
+ "learning_rate": 0.0002523304857944206,
1150
+ "loss": 0.0946,
1151
+ "step": 69000
1152
+ },
1153
+ {
1154
+ "epoch": 4.463998972316784,
1155
+ "grad_norm": 0.5030530095100403,
1156
+ "learning_rate": 0.00025125998244374505,
1157
+ "loss": 0.0961,
1158
+ "step": 69500
1159
+ },
1160
+ {
1161
+ "epoch": 4.496114072837048,
1162
+ "grad_norm": 0.5840058922767639,
1163
+ "learning_rate": 0.0002501894790930696,
1164
+ "loss": 0.0944,
1165
+ "step": 70000
1166
+ },
1167
+ {
1168
+ "epoch": 4.528229173357312,
1169
+ "grad_norm": 0.5052670836448669,
1170
+ "learning_rate": 0.00024912111674909546,
1171
+ "loss": 0.0956,
1172
+ "step": 70500
1173
+ },
1174
+ {
1175
+ "epoch": 4.560344273877577,
1176
+ "grad_norm": 0.4955575466156006,
1177
+ "learning_rate": 0.00024805061339841993,
1178
+ "loss": 0.0949,
1179
+ "step": 71000
1180
+ },
1181
+ {
1182
+ "epoch": 4.592459374397842,
1183
+ "grad_norm": 0.43502339720726013,
1184
+ "learning_rate": 0.00024698011004774446,
1185
+ "loss": 0.0939,
1186
+ "step": 71500
1187
+ },
1188
+ {
1189
+ "epoch": 4.600295458924786,
1190
+ "eval_loss": 0.09654521197080612,
1191
+ "eval_runtime": 7.343,
1192
+ "eval_samples_per_second": 68.092,
1193
+ "eval_steps_per_second": 8.58,
1194
+ "step": 71622
1195
+ },
1196
+ {
1197
+ "epoch": 4.624574474918107,
1198
+ "grad_norm": 0.3835175335407257,
1199
+ "learning_rate": 0.000245909606697069,
1200
+ "loss": 0.0968,
1201
+ "step": 72000
1202
+ },
1203
+ {
1204
+ "epoch": 4.656689575438371,
1205
+ "grad_norm": 0.46231910586357117,
1206
+ "learning_rate": 0.00024483910334639346,
1207
+ "loss": 0.0943,
1208
+ "step": 72500
1209
+ },
1210
+ {
1211
+ "epoch": 4.688804675958636,
1212
+ "grad_norm": 0.4383336901664734,
1213
+ "learning_rate": 0.00024377074100241934,
1214
+ "loss": 0.0947,
1215
+ "step": 73000
1216
+ },
1217
+ {
1218
+ "epoch": 4.7209197764789,
1219
+ "grad_norm": 0.5691344141960144,
1220
+ "learning_rate": 0.00024270023765174387,
1221
+ "loss": 0.0934,
1222
+ "step": 73500
1223
+ },
1224
+ {
1225
+ "epoch": 4.753034876999165,
1226
+ "grad_norm": 0.4115155041217804,
1227
+ "learning_rate": 0.00024162973430106837,
1228
+ "loss": 0.0935,
1229
+ "step": 74000
1230
+ },
1231
+ {
1232
+ "epoch": 4.7851499775194295,
1233
+ "grad_norm": 0.6333282589912415,
1234
+ "learning_rate": 0.00024055923095039287,
1235
+ "loss": 0.0938,
1236
+ "step": 74500
1237
+ },
1238
+ {
1239
+ "epoch": 4.800308304964995,
1240
+ "eval_loss": 0.09797964990139008,
1241
+ "eval_runtime": 7.348,
1242
+ "eval_samples_per_second": 68.046,
1243
+ "eval_steps_per_second": 8.574,
1244
+ "step": 74736
1245
+ },
1246
+ {
1247
+ "epoch": 4.817265078039695,
1248
+ "grad_norm": 0.6069188117980957,
1249
+ "learning_rate": 0.0002394887275997174,
1250
+ "loss": 0.0938,
1251
+ "step": 75000
1252
+ },
1253
+ {
1254
+ "epoch": 4.849380178559959,
1255
+ "grad_norm": 0.48266521096229553,
1256
+ "learning_rate": 0.0002384182242490419,
1257
+ "loss": 0.092,
1258
+ "step": 75500
1259
+ },
1260
+ {
1261
+ "epoch": 4.881495279080223,
1262
+ "grad_norm": 0.5051959156990051,
1263
+ "learning_rate": 0.00023734772089836642,
1264
+ "loss": 0.0938,
1265
+ "step": 76000
1266
+ },
1267
+ {
1268
+ "epoch": 4.913610379600488,
1269
+ "grad_norm": 0.36517608165740967,
1270
+ "learning_rate": 0.00023627721754769092,
1271
+ "loss": 0.0963,
1272
+ "step": 76500
1273
+ },
1274
+ {
1275
+ "epoch": 4.945725480120752,
1276
+ "grad_norm": 0.5914227366447449,
1277
+ "learning_rate": 0.00023520671419701542,
1278
+ "loss": 0.0941,
1279
+ "step": 77000
1280
+ },
1281
+ {
1282
+ "epoch": 4.9778405806410175,
1283
+ "grad_norm": 0.44252389669418335,
1284
+ "learning_rate": 0.00023414049285974263,
1285
+ "loss": 0.093,
1286
+ "step": 77500
1287
+ },
1288
+ {
1289
+ "epoch": 5.000321151005203,
1290
+ "eval_loss": 0.09499984979629517,
1291
+ "eval_runtime": 7.5722,
1292
+ "eval_samples_per_second": 66.031,
1293
+ "eval_steps_per_second": 8.32,
1294
+ "step": 77850
1295
+ },
1296
+ {
1297
+ "epoch": 5.009955681161282,
1298
+ "grad_norm": 0.501140296459198,
1299
+ "learning_rate": 0.00023306998950906716,
1300
+ "loss": 0.0918,
1301
+ "step": 78000
1302
+ },
1303
+ {
1304
+ "epoch": 5.042070781681547,
1305
+ "grad_norm": 0.3370555639266968,
1306
+ "learning_rate": 0.00023199948615839166,
1307
+ "loss": 0.0856,
1308
+ "step": 78500
1309
+ },
1310
+ {
1311
+ "epoch": 5.074185882201811,
1312
+ "grad_norm": 0.48829546570777893,
1313
+ "learning_rate": 0.00023092898280771621,
1314
+ "loss": 0.0839,
1315
+ "step": 79000
1316
+ },
1317
+ {
1318
+ "epoch": 5.106300982722076,
1319
+ "grad_norm": 0.8453528881072998,
1320
+ "learning_rate": 0.00022985847945704074,
1321
+ "loss": 0.0844,
1322
+ "step": 79500
1323
+ },
1324
+ {
1325
+ "epoch": 5.1384160832423404,
1326
+ "grad_norm": 0.4241456687450409,
1327
+ "learning_rate": 0.00022878797610636524,
1328
+ "loss": 0.0876,
1329
+ "step": 80000
1330
+ },
1331
+ {
1332
+ "epoch": 5.170531183762606,
1333
+ "grad_norm": 0.36959952116012573,
1334
+ "learning_rate": 0.00022771747275568974,
1335
+ "loss": 0.0858,
1336
+ "step": 80500
1337
+ },
1338
+ {
1339
+ "epoch": 5.20033399704541,
1340
+ "eval_loss": 0.0946323573589325,
1341
+ "eval_runtime": 7.3392,
1342
+ "eval_samples_per_second": 68.127,
1343
+ "eval_steps_per_second": 8.584,
1344
+ "step": 80964
1345
+ },
1346
+ {
1347
+ "epoch": 5.20264628428287,
1348
+ "grad_norm": 0.39433231949806213,
1349
+ "learning_rate": 0.00022664911041171562,
1350
+ "loss": 0.0864,
1351
+ "step": 81000
1352
+ },
1353
+ {
1354
+ "epoch": 5.234761384803134,
1355
+ "grad_norm": 0.724213719367981,
1356
+ "learning_rate": 0.00022557860706104012,
1357
+ "loss": 0.0856,
1358
+ "step": 81500
1359
+ },
1360
+ {
1361
+ "epoch": 5.266876485323399,
1362
+ "grad_norm": 0.3246055841445923,
1363
+ "learning_rate": 0.00022450810371036465,
1364
+ "loss": 0.0865,
1365
+ "step": 82000
1366
+ },
1367
+ {
1368
+ "epoch": 5.298991585843663,
1369
+ "grad_norm": 0.6283591389656067,
1370
+ "learning_rate": 0.00022343760035968915,
1371
+ "loss": 0.0871,
1372
+ "step": 82500
1373
+ },
1374
+ {
1375
+ "epoch": 5.3311066863639285,
1376
+ "grad_norm": 0.6402297616004944,
1377
+ "learning_rate": 0.00022236709700901365,
1378
+ "loss": 0.0853,
1379
+ "step": 83000
1380
+ },
1381
+ {
1382
+ "epoch": 5.363221786884193,
1383
+ "grad_norm": 0.4186316728591919,
1384
+ "learning_rate": 0.00022129873466503953,
1385
+ "loss": 0.0879,
1386
+ "step": 83500
1387
+ },
1388
+ {
1389
+ "epoch": 5.395336887404458,
1390
+ "grad_norm": 0.39480188488960266,
1391
+ "learning_rate": 0.00022022823131436403,
1392
+ "loss": 0.0881,
1393
+ "step": 84000
1394
+ },
1395
+ {
1396
+ "epoch": 5.400346843085619,
1397
+ "eval_loss": 0.09650416672229767,
1398
+ "eval_runtime": 7.3429,
1399
+ "eval_samples_per_second": 68.093,
1400
+ "eval_steps_per_second": 8.58,
1401
+ "step": 84078
1402
+ },
1403
+ {
1404
+ "epoch": 5.427451987924722,
1405
+ "grad_norm": 0.45581161975860596,
1406
+ "learning_rate": 0.00021915772796368853,
1407
+ "loss": 0.0876,
1408
+ "step": 84500
1409
+ },
1410
+ {
1411
+ "epoch": 5.459567088444987,
1412
+ "grad_norm": 0.5241764187812805,
1413
+ "learning_rate": 0.00021808722461301306,
1414
+ "loss": 0.0874,
1415
+ "step": 85000
1416
+ },
1417
+ {
1418
+ "epoch": 5.491682188965251,
1419
+ "grad_norm": 0.5250868797302246,
1420
+ "learning_rate": 0.00021701672126233756,
1421
+ "loss": 0.0875,
1422
+ "step": 85500
1423
+ },
1424
+ {
1425
+ "epoch": 5.5237972894855165,
1426
+ "grad_norm": 0.4289800226688385,
1427
+ "learning_rate": 0.00021594621791166209,
1428
+ "loss": 0.0861,
1429
+ "step": 86000
1430
+ },
1431
+ {
1432
+ "epoch": 5.555912390005781,
1433
+ "grad_norm": 0.4157341718673706,
1434
+ "learning_rate": 0.00021487571456098659,
1435
+ "loss": 0.0845,
1436
+ "step": 86500
1437
+ },
1438
+ {
1439
+ "epoch": 5.588027490526045,
1440
+ "grad_norm": 0.8754804730415344,
1441
+ "learning_rate": 0.00021380521121031109,
1442
+ "loss": 0.0883,
1443
+ "step": 87000
1444
+ },
1445
+ {
1446
+ "epoch": 5.600359689125827,
1447
+ "eval_loss": 0.09484092891216278,
1448
+ "eval_runtime": 7.3492,
1449
+ "eval_samples_per_second": 68.034,
1450
+ "eval_steps_per_second": 8.572,
1451
+ "step": 87192
1452
+ },
1453
+ {
1454
+ "epoch": 5.62014259104631,
1455
+ "grad_norm": 0.5359424352645874,
1456
+ "learning_rate": 0.0002127347078596356,
1457
+ "loss": 0.0853,
1458
+ "step": 87500
1459
+ },
1460
+ {
1461
+ "epoch": 5.652257691566574,
1462
+ "grad_norm": 0.44357284903526306,
1463
+ "learning_rate": 0.00021166848652236282,
1464
+ "loss": 0.0868,
1465
+ "step": 88000
1466
+ },
1467
+ {
1468
+ "epoch": 5.684372792086839,
1469
+ "grad_norm": 0.5361419916152954,
1470
+ "learning_rate": 0.00021059798317168732,
1471
+ "loss": 0.0877,
1472
+ "step": 88500
1473
+ },
1474
+ {
1475
+ "epoch": 5.716487892607104,
1476
+ "grad_norm": 0.29155826568603516,
1477
+ "learning_rate": 0.00020952747982101185,
1478
+ "loss": 0.0863,
1479
+ "step": 89000
1480
+ },
1481
+ {
1482
+ "epoch": 5.748602993127369,
1483
+ "grad_norm": 0.5860145688056946,
1484
+ "learning_rate": 0.00020845697647033635,
1485
+ "loss": 0.0874,
1486
+ "step": 89500
1487
+ },
1488
+ {
1489
+ "epoch": 5.780718093647633,
1490
+ "grad_norm": 0.5508409738540649,
1491
+ "learning_rate": 0.00020738647311966085,
1492
+ "loss": 0.0846,
1493
+ "step": 90000
1494
+ },
1495
+ {
1496
+ "epoch": 5.800372535166035,
1497
+ "eval_loss": 0.0940055400133133,
1498
+ "eval_runtime": 7.3511,
1499
+ "eval_samples_per_second": 68.017,
1500
+ "eval_steps_per_second": 8.57,
1501
+ "step": 90306
1502
+ },
1503
+ {
1504
+ "epoch": 5.812833194167898,
1505
+ "grad_norm": 0.5767956376075745,
1506
+ "learning_rate": 0.00020631596976898538,
1507
+ "loss": 0.0845,
1508
+ "step": 90500
1509
+ },
1510
+ {
1511
+ "epoch": 5.844948294688162,
1512
+ "grad_norm": 0.45675429701805115,
1513
+ "learning_rate": 0.00020524546641830988,
1514
+ "loss": 0.0859,
1515
+ "step": 91000
1516
+ },
1517
+ {
1518
+ "epoch": 5.8770633952084275,
1519
+ "grad_norm": 0.4499921500682831,
1520
+ "learning_rate": 0.0002041749630676344,
1521
+ "loss": 0.0882,
1522
+ "step": 91500
1523
+ },
1524
+ {
1525
+ "epoch": 5.909178495728692,
1526
+ "grad_norm": 0.4361652731895447,
1527
+ "learning_rate": 0.0002031044597169589,
1528
+ "loss": 0.088,
1529
+ "step": 92000
1530
+ },
1531
+ {
1532
+ "epoch": 5.941293596248956,
1533
+ "grad_norm": 0.46764621138572693,
1534
+ "learning_rate": 0.00020203395636628343,
1535
+ "loss": 0.0877,
1536
+ "step": 92500
1537
+ },
1538
+ {
1539
+ "epoch": 5.973408696769221,
1540
+ "grad_norm": 0.46029818058013916,
1541
+ "learning_rate": 0.00020096559402230929,
1542
+ "loss": 0.0862,
1543
+ "step": 93000
1544
+ },
1545
+ {
1546
+ "epoch": 6.000385381206243,
1547
+ "eval_loss": 0.09267408400774002,
1548
+ "eval_runtime": 7.5179,
1549
+ "eval_samples_per_second": 66.508,
1550
+ "eval_steps_per_second": 8.38,
1551
+ "step": 93420
1552
+ },
1553
+ {
1554
+ "epoch": 6.005523797289485,
1555
+ "grad_norm": 0.37162384390830994,
1556
+ "learning_rate": 0.0001998950906716338,
1557
+ "loss": 0.0849,
1558
+ "step": 93500
1559
+ },
1560
+ {
1561
+ "epoch": 6.03763889780975,
1562
+ "grad_norm": 0.60701984167099,
1563
+ "learning_rate": 0.0001988245873209583,
1564
+ "loss": 0.0765,
1565
+ "step": 94000
1566
+ },
1567
+ {
1568
+ "epoch": 6.069753998330015,
1569
+ "grad_norm": 0.41340622305870056,
1570
+ "learning_rate": 0.0001977562249769842,
1571
+ "loss": 0.0794,
1572
+ "step": 94500
1573
+ },
1574
+ {
1575
+ "epoch": 6.10186909885028,
1576
+ "grad_norm": 0.4631298780441284,
1577
+ "learning_rate": 0.0001966857216263087,
1578
+ "loss": 0.0775,
1579
+ "step": 95000
1580
+ },
1581
+ {
1582
+ "epoch": 6.133984199370544,
1583
+ "grad_norm": 0.4848615527153015,
1584
+ "learning_rate": 0.0001956152182756332,
1585
+ "loss": 0.0772,
1586
+ "step": 95500
1587
+ },
1588
+ {
1589
+ "epoch": 6.166099299890809,
1590
+ "grad_norm": 0.4167845845222473,
1591
+ "learning_rate": 0.00019454471492495772,
1592
+ "loss": 0.0794,
1593
+ "step": 96000
1594
+ },
1595
+ {
1596
+ "epoch": 6.198214400411073,
1597
+ "grad_norm": 0.48452332615852356,
1598
+ "learning_rate": 0.00019347421157428222,
1599
+ "loss": 0.0786,
1600
+ "step": 96500
1601
+ },
1602
+ {
1603
+ "epoch": 6.200398227246451,
1604
+ "eval_loss": 0.09262655675411224,
1605
+ "eval_runtime": 7.3483,
1606
+ "eval_samples_per_second": 68.043,
1607
+ "eval_steps_per_second": 8.573,
1608
+ "step": 96534
1609
+ },
1610
+ {
1611
+ "epoch": 6.2303295009313375,
1612
+ "grad_norm": 0.5630140900611877,
1613
+ "learning_rate": 0.00019240370822360675,
1614
+ "loss": 0.08,
1615
+ "step": 97000
1616
+ },
1617
+ {
1618
+ "epoch": 6.262444601451603,
1619
+ "grad_norm": 0.6407498717308044,
1620
+ "learning_rate": 0.00019133320487293128,
1621
+ "loss": 0.0772,
1622
+ "step": 97500
1623
+ },
1624
+ {
1625
+ "epoch": 6.294559701971867,
1626
+ "grad_norm": 0.2974085211753845,
1627
+ "learning_rate": 0.00019026270152225577,
1628
+ "loss": 0.0768,
1629
+ "step": 98000
1630
+ },
1631
+ {
1632
+ "epoch": 6.326674802492132,
1633
+ "grad_norm": 0.4953305721282959,
1634
+ "learning_rate": 0.0001891921981715803,
1635
+ "loss": 0.0794,
1636
+ "step": 98500
1637
+ },
1638
+ {
1639
+ "epoch": 6.358789903012396,
1640
+ "grad_norm": 0.5464000105857849,
1641
+ "learning_rate": 0.00018812383582760616,
1642
+ "loss": 0.0791,
1643
+ "step": 99000
1644
+ },
1645
+ {
1646
+ "epoch": 6.390905003532661,
1647
+ "grad_norm": 0.47236204147338867,
1648
+ "learning_rate": 0.00018705333247693066,
1649
+ "loss": 0.0795,
1650
+ "step": 99500
1651
+ },
1652
+ {
1653
+ "epoch": 6.4004110732866595,
1654
+ "eval_loss": 0.09152522683143616,
1655
+ "eval_runtime": 7.3533,
1656
+ "eval_samples_per_second": 67.997,
1657
+ "eval_steps_per_second": 8.568,
1658
+ "step": 99648
1659
+ },
1660
+ {
1661
+ "epoch": 6.423020104052926,
1662
+ "grad_norm": 0.5907355546951294,
1663
+ "learning_rate": 0.00018598282912625518,
1664
+ "loss": 0.0788,
1665
+ "step": 100000
1666
+ },
1667
+ {
1668
+ "epoch": 6.455135204573191,
1669
+ "grad_norm": 0.4848667085170746,
1670
+ "learning_rate": 0.00018491232577557968,
1671
+ "loss": 0.0793,
1672
+ "step": 100500
1673
+ },
1674
+ {
1675
+ "epoch": 6.487250305093455,
1676
+ "grad_norm": 0.5288995504379272,
1677
+ "learning_rate": 0.0001838418224249042,
1678
+ "loss": 0.0787,
1679
+ "step": 101000
1680
+ },
1681
+ {
1682
+ "epoch": 6.519365405613719,
1683
+ "grad_norm": 0.3318091034889221,
1684
+ "learning_rate": 0.0001827713190742287,
1685
+ "loss": 0.0774,
1686
+ "step": 101500
1687
+ },
1688
+ {
1689
+ "epoch": 6.551480506133984,
1690
+ "grad_norm": 0.5783077478408813,
1691
+ "learning_rate": 0.0001817008157235532,
1692
+ "loss": 0.0786,
1693
+ "step": 102000
1694
+ },
1695
+ {
1696
+ "epoch": 6.5835956066542485,
1697
+ "grad_norm": 0.748836874961853,
1698
+ "learning_rate": 0.0001806324533795791,
1699
+ "loss": 0.0811,
1700
+ "step": 102500
1701
+ },
1702
+ {
1703
+ "epoch": 6.600423919326867,
1704
+ "eval_loss": 0.09013129770755768,
1705
+ "eval_runtime": 7.3443,
1706
+ "eval_samples_per_second": 68.08,
1707
+ "eval_steps_per_second": 8.578,
1708
+ "step": 102762
1709
+ },
1710
+ {
1711
+ "epoch": 6.615710707174514,
1712
+ "grad_norm": 0.4239063262939453,
1713
+ "learning_rate": 0.00017956409103560495,
1714
+ "loss": 0.0779,
1715
+ "step": 103000
1716
+ },
1717
+ {
1718
+ "epoch": 6.647825807694778,
1719
+ "grad_norm": 0.42676037549972534,
1720
+ "learning_rate": 0.00017849358768492945,
1721
+ "loss": 0.0801,
1722
+ "step": 103500
1723
+ },
1724
+ {
1725
+ "epoch": 6.679940908215043,
1726
+ "grad_norm": 0.4551810026168823,
1727
+ "learning_rate": 0.00017742308433425397,
1728
+ "loss": 0.0787,
1729
+ "step": 104000
1730
+ },
1731
+ {
1732
+ "epoch": 6.712056008735307,
1733
+ "grad_norm": 0.8408447504043579,
1734
+ "learning_rate": 0.0001763525809835785,
1735
+ "loss": 0.077,
1736
+ "step": 104500
1737
+ },
1738
+ {
1739
+ "epoch": 6.744171109255572,
1740
+ "grad_norm": 0.4141174554824829,
1741
+ "learning_rate": 0.000175282077632903,
1742
+ "loss": 0.0774,
1743
+ "step": 105000
1744
+ },
1745
+ {
1746
+ "epoch": 6.7762862097758365,
1747
+ "grad_norm": 0.4172053933143616,
1748
+ "learning_rate": 0.00017421157428222753,
1749
+ "loss": 0.0783,
1750
+ "step": 105500
1751
+ },
1752
+ {
1753
+ "epoch": 6.800436765367076,
1754
+ "eval_loss": 0.08964564651250839,
1755
+ "eval_runtime": 7.3548,
1756
+ "eval_samples_per_second": 67.983,
1757
+ "eval_steps_per_second": 8.566,
1758
+ "step": 105876
1759
+ },
1760
+ {
1761
+ "epoch": 6.808401310296102,
1762
+ "grad_norm": 0.4714060127735138,
1763
+ "learning_rate": 0.00017314107093155203,
1764
+ "loss": 0.079,
1765
+ "step": 106000
1766
+ },
1767
+ {
1768
+ "epoch": 6.840516410816366,
1769
+ "grad_norm": 0.5461053252220154,
1770
+ "learning_rate": 0.00017207056758087653,
1771
+ "loss": 0.0776,
1772
+ "step": 106500
1773
+ },
1774
+ {
1775
+ "epoch": 6.87263151133663,
1776
+ "grad_norm": 0.32436585426330566,
1777
+ "learning_rate": 0.00017100006423020106,
1778
+ "loss": 0.0807,
1779
+ "step": 107000
1780
+ },
1781
+ {
1782
+ "epoch": 6.904746611856895,
1783
+ "grad_norm": 0.46811825037002563,
1784
+ "learning_rate": 0.00016992956087952556,
1785
+ "loss": 0.0779,
1786
+ "step": 107500
1787
+ },
1788
+ {
1789
+ "epoch": 6.936861712377159,
1790
+ "grad_norm": 0.5478762984275818,
1791
+ "learning_rate": 0.0001688611985355514,
1792
+ "loss": 0.0762,
1793
+ "step": 108000
1794
+ },
1795
+ {
1796
+ "epoch": 6.9689768128974245,
1797
+ "grad_norm": 0.4181552529335022,
1798
+ "learning_rate": 0.00016779069518487594,
1799
+ "loss": 0.0803,
1800
+ "step": 108500
1801
+ },
1802
+ {
1803
+ "epoch": 7.000449611407284,
1804
+ "eval_loss": 0.08895646780729294,
1805
+ "eval_runtime": 7.474,
1806
+ "eval_samples_per_second": 66.899,
1807
+ "eval_steps_per_second": 8.429,
1808
+ "step": 108990
1809
+ },
1810
+ {
1811
+ "epoch": 7.001091913417689,
1812
+ "grad_norm": 0.32527291774749756,
1813
+ "learning_rate": 0.00016672019183420044,
1814
+ "loss": 0.0779,
1815
+ "step": 109000
1816
+ },
1817
+ {
1818
+ "epoch": 7.033207013937954,
1819
+ "grad_norm": 0.535256564617157,
1820
+ "learning_rate": 0.00016565182949022632,
1821
+ "loss": 0.0683,
1822
+ "step": 109500
1823
+ },
1824
+ {
1825
+ "epoch": 7.065322114458218,
1826
+ "grad_norm": 0.5849858522415161,
1827
+ "learning_rate": 0.00016458132613955082,
1828
+ "loss": 0.0701,
1829
+ "step": 110000
1830
+ },
1831
+ {
1832
+ "epoch": 7.097437214978483,
1833
+ "grad_norm": 0.3674163222312927,
1834
+ "learning_rate": 0.00016351082278887532,
1835
+ "loss": 0.0709,
1836
+ "step": 110500
1837
+ },
1838
+ {
1839
+ "epoch": 7.1295523154987475,
1840
+ "grad_norm": 0.44683510065078735,
1841
+ "learning_rate": 0.00016244031943819985,
1842
+ "loss": 0.0707,
1843
+ "step": 111000
1844
+ },
1845
+ {
1846
+ "epoch": 7.161667416019013,
1847
+ "grad_norm": 0.5484521985054016,
1848
+ "learning_rate": 0.00016136981608752437,
1849
+ "loss": 0.0717,
1850
+ "step": 111500
1851
+ },
1852
+ {
1853
+ "epoch": 7.193782516539277,
1854
+ "grad_norm": 0.3398258686065674,
1855
+ "learning_rate": 0.00016030145374355023,
1856
+ "loss": 0.0721,
1857
+ "step": 112000
1858
+ },
1859
+ {
1860
+ "epoch": 7.200462457447491,
1861
+ "eval_loss": 0.08926703035831451,
1862
+ "eval_runtime": 7.3479,
1863
+ "eval_samples_per_second": 68.047,
1864
+ "eval_steps_per_second": 8.574,
1865
+ "step": 112104
1866
+ },
1867
+ {
1868
+ "epoch": 7.225897617059541,
1869
+ "grad_norm": 0.5869883298873901,
1870
+ "learning_rate": 0.00015923095039287476,
1871
+ "loss": 0.0695,
1872
+ "step": 112500
1873
+ },
1874
+ {
1875
+ "epoch": 7.258012717579806,
1876
+ "grad_norm": 0.38123539090156555,
1877
+ "learning_rate": 0.00015816044704219926,
1878
+ "loss": 0.0702,
1879
+ "step": 113000
1880
+ },
1881
+ {
1882
+ "epoch": 7.29012781810007,
1883
+ "grad_norm": 0.4954741299152374,
1884
+ "learning_rate": 0.00015708994369152376,
1885
+ "loss": 0.0708,
1886
+ "step": 113500
1887
+ },
1888
+ {
1889
+ "epoch": 7.3222429186203355,
1890
+ "grad_norm": 0.3034817576408386,
1891
+ "learning_rate": 0.00015601944034084828,
1892
+ "loss": 0.0719,
1893
+ "step": 114000
1894
+ },
1895
+ {
1896
+ "epoch": 7.3543580191406,
1897
+ "grad_norm": 0.44598105549812317,
1898
+ "learning_rate": 0.00015494893699017278,
1899
+ "loss": 0.0699,
1900
+ "step": 114500
1901
+ },
1902
+ {
1903
+ "epoch": 7.386473119660865,
1904
+ "grad_norm": 0.4353293776512146,
1905
+ "learning_rate": 0.0001538784336394973,
1906
+ "loss": 0.0707,
1907
+ "step": 115000
1908
+ },
1909
+ {
1910
+ "epoch": 7.4004753034877,
1911
+ "eval_loss": 0.08555440604686737,
1912
+ "eval_runtime": 7.3485,
1913
+ "eval_samples_per_second": 68.041,
1914
+ "eval_steps_per_second": 8.573,
1915
+ "step": 115218
1916
+ },
1917
+ {
1918
+ "epoch": 7.418588220181129,
1919
+ "grad_norm": 0.4826795160770416,
1920
+ "learning_rate": 0.0001528079302888218,
1921
+ "loss": 0.0711,
1922
+ "step": 115500
1923
+ },
1924
+ {
1925
+ "epoch": 7.450703320701394,
1926
+ "grad_norm": 0.33973926305770874,
1927
+ "learning_rate": 0.00015173956794484766,
1928
+ "loss": 0.0699,
1929
+ "step": 116000
1930
+ },
1931
+ {
1932
+ "epoch": 7.482818421221658,
1933
+ "grad_norm": 0.5806561708450317,
1934
+ "learning_rate": 0.0001506690645941722,
1935
+ "loss": 0.0721,
1936
+ "step": 116500
1937
+ },
1938
+ {
1939
+ "epoch": 7.5149335217419235,
1940
+ "grad_norm": 0.45582470297813416,
1941
+ "learning_rate": 0.0001495985612434967,
1942
+ "loss": 0.0706,
1943
+ "step": 117000
1944
+ },
1945
+ {
1946
+ "epoch": 7.547048622262188,
1947
+ "grad_norm": 0.40795230865478516,
1948
+ "learning_rate": 0.0001485280578928212,
1949
+ "loss": 0.0699,
1950
+ "step": 117500
1951
+ },
1952
+ {
1953
+ "epoch": 7.579163722782452,
1954
+ "grad_norm": 0.5874589681625366,
1955
+ "learning_rate": 0.00014745755454214572,
1956
+ "loss": 0.0721,
1957
+ "step": 118000
1958
+ },
1959
+ {
1960
+ "epoch": 7.600488149527908,
1961
+ "eval_loss": 0.08731518685817719,
1962
+ "eval_runtime": 7.3538,
1963
+ "eval_samples_per_second": 67.992,
1964
+ "eval_steps_per_second": 8.567,
1965
+ "step": 118332
1966
+ },
1967
+ {
1968
+ "epoch": 7.611278823302717,
1969
+ "grad_norm": 0.5102195739746094,
1970
+ "learning_rate": 0.00014638705119147022,
1971
+ "loss": 0.0723,
1972
+ "step": 118500
1973
+ },
1974
+ {
1975
+ "epoch": 7.643393923822981,
1976
+ "grad_norm": 0.8326733708381653,
1977
+ "learning_rate": 0.0001453186888474961,
1978
+ "loss": 0.0707,
1979
+ "step": 119000
1980
+ },
1981
+ {
1982
+ "epoch": 7.6755090243432464,
1983
+ "grad_norm": 0.4096178710460663,
1984
+ "learning_rate": 0.00014424818549682063,
1985
+ "loss": 0.071,
1986
+ "step": 119500
1987
+ },
1988
+ {
1989
+ "epoch": 7.707624124863511,
1990
+ "grad_norm": 0.4716354310512543,
1991
+ "learning_rate": 0.00014317768214614513,
1992
+ "loss": 0.0702,
1993
+ "step": 120000
1994
+ },
1995
+ {
1996
+ "epoch": 7.739739225383776,
1997
+ "grad_norm": 0.3855126202106476,
1998
+ "learning_rate": 0.00014210717879546965,
1999
+ "loss": 0.0684,
2000
+ "step": 120500
2001
+ },
2002
+ {
2003
+ "epoch": 7.77185432590404,
2004
+ "grad_norm": 0.6053977012634277,
2005
+ "learning_rate": 0.0001410388164514955,
2006
+ "loss": 0.0706,
2007
+ "step": 121000
2008
+ },
2009
+ {
2010
+ "epoch": 7.800500995568116,
2011
+ "eval_loss": 0.08562512695789337,
2012
+ "eval_runtime": 7.3498,
2013
+ "eval_samples_per_second": 68.029,
2014
+ "eval_steps_per_second": 8.572,
2015
+ "step": 121446
2016
+ },
2017
+ {
2018
+ "epoch": 7.803969426424304,
2019
+ "grad_norm": 0.34569281339645386,
2020
+ "learning_rate": 0.00013996831310082,
2021
+ "loss": 0.07,
2022
+ "step": 121500
2023
+ },
2024
+ {
2025
+ "epoch": 7.836084526944569,
2026
+ "grad_norm": 0.5277242064476013,
2027
+ "learning_rate": 0.00013889780975014454,
2028
+ "loss": 0.0707,
2029
+ "step": 122000
2030
+ },
2031
+ {
2032
+ "epoch": 7.868199627464834,
2033
+ "grad_norm": 0.49101319909095764,
2034
+ "learning_rate": 0.00013782730639946904,
2035
+ "loss": 0.072,
2036
+ "step": 122500
2037
+ },
2038
+ {
2039
+ "epoch": 7.900314727985099,
2040
+ "grad_norm": 0.612022340297699,
2041
+ "learning_rate": 0.00013675680304879354,
2042
+ "loss": 0.0718,
2043
+ "step": 123000
2044
+ },
2045
+ {
2046
+ "epoch": 7.932429828505363,
2047
+ "grad_norm": 0.43211016058921814,
2048
+ "learning_rate": 0.00013568844070481942,
2049
+ "loss": 0.0703,
2050
+ "step": 123500
2051
+ },
2052
+ {
2053
+ "epoch": 7.964544929025628,
2054
+ "grad_norm": 0.5008595585823059,
2055
+ "learning_rate": 0.00013461793735414392,
2056
+ "loss": 0.0711,
2057
+ "step": 124000
2058
+ },
2059
+ {
2060
+ "epoch": 7.996660029545892,
2061
+ "grad_norm": 0.42193835973739624,
2062
+ "learning_rate": 0.00013354957501016977,
2063
+ "loss": 0.0705,
2064
+ "step": 124500
2065
+ },
2066
+ {
2067
+ "epoch": 8.000513841608324,
2068
+ "eval_loss": 0.0847671627998352,
2069
+ "eval_runtime": 7.4721,
2070
+ "eval_samples_per_second": 66.916,
2071
+ "eval_steps_per_second": 8.431,
2072
+ "step": 124560
2073
+ },
2074
+ {
2075
+ "epoch": 8.028775130066157,
2076
+ "grad_norm": 0.4556228816509247,
2077
+ "learning_rate": 0.0001324790716594943,
2078
+ "loss": 0.0619,
2079
+ "step": 125000
2080
+ },
2081
+ {
2082
+ "epoch": 8.060890230586422,
2083
+ "grad_norm": 0.41463834047317505,
2084
+ "learning_rate": 0.0001314085683088188,
2085
+ "loss": 0.0607,
2086
+ "step": 125500
2087
+ },
2088
+ {
2089
+ "epoch": 8.093005331106687,
2090
+ "grad_norm": 0.6098557710647583,
2091
+ "learning_rate": 0.00013033806495814333,
2092
+ "loss": 0.0629,
2093
+ "step": 126000
2094
+ },
2095
+ {
2096
+ "epoch": 8.125120431626952,
2097
+ "grad_norm": 0.4417126774787903,
2098
+ "learning_rate": 0.00012926756160746785,
2099
+ "loss": 0.0621,
2100
+ "step": 126500
2101
+ },
2102
+ {
2103
+ "epoch": 8.157235532147215,
2104
+ "grad_norm": 1.175355076789856,
2105
+ "learning_rate": 0.00012819705825679235,
2106
+ "loss": 0.0617,
2107
+ "step": 127000
2108
+ },
2109
+ {
2110
+ "epoch": 8.18935063266748,
2111
+ "grad_norm": 0.530025064945221,
2112
+ "learning_rate": 0.00012712655490611688,
2113
+ "loss": 0.0634,
2114
+ "step": 127500
2115
+ },
2116
+ {
2117
+ "epoch": 8.200526687648532,
2118
+ "eval_loss": 0.08836065232753754,
2119
+ "eval_runtime": 7.3681,
2120
+ "eval_samples_per_second": 67.86,
2121
+ "eval_steps_per_second": 8.55,
2122
+ "step": 127674
2123
+ },
2124
+ {
2125
+ "epoch": 8.221465733187745,
2126
+ "grad_norm": 0.5124799013137817,
2127
+ "learning_rate": 0.00012605605155544138,
2128
+ "loss": 0.0634,
2129
+ "step": 128000
2130
+ },
2131
+ {
2132
+ "epoch": 8.253580833708009,
2133
+ "grad_norm": 0.6200098395347595,
2134
+ "learning_rate": 0.00012498554820476588,
2135
+ "loss": 0.0628,
2136
+ "step": 128500
2137
+ },
2138
+ {
2139
+ "epoch": 8.285695934228274,
2140
+ "grad_norm": 0.403604656457901,
2141
+ "learning_rate": 0.0001239150448540904,
2142
+ "loss": 0.0633,
2143
+ "step": 129000
2144
+ },
2145
+ {
2146
+ "epoch": 8.317811034748539,
2147
+ "grad_norm": 0.5007002949714661,
2148
+ "learning_rate": 0.00012284668251011626,
2149
+ "loss": 0.0636,
2150
+ "step": 129500
2151
+ },
2152
+ {
2153
+ "epoch": 8.349926135268804,
2154
+ "grad_norm": 0.5700250864028931,
2155
+ "learning_rate": 0.00012177617915944078,
2156
+ "loss": 0.0639,
2157
+ "step": 130000
2158
+ },
2159
+ {
2160
+ "epoch": 8.382041235789067,
2161
+ "grad_norm": 0.46153202652931213,
2162
+ "learning_rate": 0.00012070567580876529,
2163
+ "loss": 0.0632,
2164
+ "step": 130500
2165
+ },
2166
+ {
2167
+ "epoch": 8.40053953368874,
2168
+ "eval_loss": 0.08482711762189865,
2169
+ "eval_runtime": 7.3637,
2170
+ "eval_samples_per_second": 67.901,
2171
+ "eval_steps_per_second": 8.556,
2172
+ "step": 130788
2173
+ },
2174
+ {
2175
+ "epoch": 8.414156336309333,
2176
+ "grad_norm": 0.4853110611438751,
2177
+ "learning_rate": 0.00011963517245808979,
2178
+ "loss": 0.0636,
2179
+ "step": 131000
2180
+ },
2181
+ {
2182
+ "epoch": 8.446271436829598,
2183
+ "grad_norm": 0.6121137142181396,
2184
+ "learning_rate": 0.0001185646691074143,
2185
+ "loss": 0.0624,
2186
+ "step": 131500
2187
+ },
2188
+ {
2189
+ "epoch": 8.478386537349863,
2190
+ "grad_norm": 0.3986789584159851,
2191
+ "learning_rate": 0.00011749416575673882,
2192
+ "loss": 0.0627,
2193
+ "step": 132000
2194
+ },
2195
+ {
2196
+ "epoch": 8.510501637870126,
2197
+ "grad_norm": 0.6468679904937744,
2198
+ "learning_rate": 0.00011642366240606333,
2199
+ "loss": 0.0618,
2200
+ "step": 132500
2201
+ },
2202
+ {
2203
+ "epoch": 8.542616738390391,
2204
+ "grad_norm": 0.49200209975242615,
2205
+ "learning_rate": 0.00011535315905538784,
2206
+ "loss": 0.0631,
2207
+ "step": 133000
2208
+ },
2209
+ {
2210
+ "epoch": 8.574731838910656,
2211
+ "grad_norm": 0.47202590107917786,
2212
+ "learning_rate": 0.00011428265570471237,
2213
+ "loss": 0.0653,
2214
+ "step": 133500
2215
+ },
2216
+ {
2217
+ "epoch": 8.60055237972895,
2218
+ "eval_loss": 0.08428142964839935,
2219
+ "eval_runtime": 7.37,
2220
+ "eval_samples_per_second": 67.843,
2221
+ "eval_steps_per_second": 8.548,
2222
+ "step": 133902
2223
+ },
2224
+ {
2225
+ "epoch": 8.60684693943092,
2226
+ "grad_norm": 0.5834782123565674,
2227
+ "learning_rate": 0.00011321429336073824,
2228
+ "loss": 0.0641,
2229
+ "step": 134000
2230
+ },
2231
+ {
2232
+ "epoch": 8.638962039951185,
2233
+ "grad_norm": 0.614308774471283,
2234
+ "learning_rate": 0.00011214379001006274,
2235
+ "loss": 0.0618,
2236
+ "step": 134500
2237
+ },
2238
+ {
2239
+ "epoch": 8.67107714047145,
2240
+ "grad_norm": 0.5445231199264526,
2241
+ "learning_rate": 0.00011107328665938725,
2242
+ "loss": 0.0636,
2243
+ "step": 135000
2244
+ },
2245
+ {
2246
+ "epoch": 8.703192240991715,
2247
+ "grad_norm": 0.5643283128738403,
2248
+ "learning_rate": 0.00011000278330871177,
2249
+ "loss": 0.062,
2250
+ "step": 135500
2251
+ },
2252
+ {
2253
+ "epoch": 8.735307341511978,
2254
+ "grad_norm": 0.5778331160545349,
2255
+ "learning_rate": 0.00010893227995803628,
2256
+ "loss": 0.0646,
2257
+ "step": 136000
2258
+ },
2259
+ {
2260
+ "epoch": 8.767422442032244,
2261
+ "grad_norm": 0.6426026821136475,
2262
+ "learning_rate": 0.00010786177660736079,
2263
+ "loss": 0.0633,
2264
+ "step": 136500
2265
+ },
2266
+ {
2267
+ "epoch": 8.799537542552509,
2268
+ "grad_norm": 0.40172818303108215,
2269
+ "learning_rate": 0.00010679127325668529,
2270
+ "loss": 0.0652,
2271
+ "step": 137000
2272
+ },
2273
+ {
2274
+ "epoch": 8.800565225769157,
2275
+ "eval_loss": 0.08410488069057465,
2276
+ "eval_runtime": 7.3612,
2277
+ "eval_samples_per_second": 67.923,
2278
+ "eval_steps_per_second": 8.558,
2279
+ "step": 137016
2280
+ },
2281
+ {
2282
+ "epoch": 8.831652643072772,
2283
+ "grad_norm": 0.5106441974639893,
2284
+ "learning_rate": 0.0001057207699060098,
2285
+ "loss": 0.0627,
2286
+ "step": 137500
2287
+ },
2288
+ {
2289
+ "epoch": 8.863767743593037,
2290
+ "grad_norm": 0.5907523036003113,
2291
+ "learning_rate": 0.00010465026655533432,
2292
+ "loss": 0.0629,
2293
+ "step": 138000
2294
+ },
2295
+ {
2296
+ "epoch": 8.895882844113302,
2297
+ "grad_norm": 0.5151248574256897,
2298
+ "learning_rate": 0.00010358190421136017,
2299
+ "loss": 0.0616,
2300
+ "step": 138500
2301
+ },
2302
+ {
2303
+ "epoch": 8.927997944633567,
2304
+ "grad_norm": 0.4213479459285736,
2305
+ "learning_rate": 0.00010251140086068469,
2306
+ "loss": 0.0641,
2307
+ "step": 139000
2308
+ },
2309
+ {
2310
+ "epoch": 8.96011304515383,
2311
+ "grad_norm": 0.5318893790245056,
2312
+ "learning_rate": 0.0001014408975100092,
2313
+ "loss": 0.0633,
2314
+ "step": 139500
2315
+ },
2316
+ {
2317
+ "epoch": 8.992228145674096,
2318
+ "grad_norm": 0.613757848739624,
2319
+ "learning_rate": 0.00010037039415933371,
2320
+ "loss": 0.0616,
2321
+ "step": 140000
2322
+ },
2323
+ {
2324
+ "epoch": 9.000578071809365,
2325
+ "eval_loss": 0.08246000856161118,
2326
+ "eval_runtime": 7.4333,
2327
+ "eval_samples_per_second": 67.265,
2328
+ "eval_steps_per_second": 8.475,
2329
+ "step": 140130
2330
+ },
2331
+ {
2332
+ "epoch": 9.024343246194361,
2333
+ "grad_norm": 0.581920862197876,
2334
+ "learning_rate": 9.929989080865824e-05,
2335
+ "loss": 0.0561,
2336
+ "step": 140500
2337
+ },
2338
+ {
2339
+ "epoch": 9.056458346714626,
2340
+ "grad_norm": 0.4592820107936859,
2341
+ "learning_rate": 9.822938745798274e-05,
2342
+ "loss": 0.0543,
2343
+ "step": 141000
2344
+ },
2345
+ {
2346
+ "epoch": 9.08857344723489,
2347
+ "grad_norm": 0.592686116695404,
2348
+ "learning_rate": 9.716102511400861e-05,
2349
+ "loss": 0.054,
2350
+ "step": 141500
2351
+ },
2352
+ {
2353
+ "epoch": 9.120688547755154,
2354
+ "grad_norm": 0.43550175428390503,
2355
+ "learning_rate": 9.609266277003447e-05,
2356
+ "loss": 0.0532,
2357
+ "step": 142000
2358
+ },
2359
+ {
2360
+ "epoch": 9.15280364827542,
2361
+ "grad_norm": 0.5796665549278259,
2362
+ "learning_rate": 9.502215941935899e-05,
2363
+ "loss": 0.0555,
2364
+ "step": 142500
2365
+ },
2366
+ {
2367
+ "epoch": 9.184918748795683,
2368
+ "grad_norm": 0.4906461834907532,
2369
+ "learning_rate": 9.39516560686835e-05,
2370
+ "loss": 0.0543,
2371
+ "step": 143000
2372
+ },
2373
+ {
2374
+ "epoch": 9.200590917849572,
2375
+ "eval_loss": 0.08360330015420914,
2376
+ "eval_runtime": 7.3448,
2377
+ "eval_samples_per_second": 68.075,
2378
+ "eval_steps_per_second": 8.577,
2379
+ "step": 143244
2380
+ },
2381
+ {
2382
+ "epoch": 9.217033849315948,
2383
+ "grad_norm": 0.37401750683784485,
2384
+ "learning_rate": 9.288329372470936e-05,
2385
+ "loss": 0.0534,
2386
+ "step": 143500
2387
+ },
2388
+ {
2389
+ "epoch": 9.249148949836213,
2390
+ "grad_norm": 0.4685318171977997,
2391
+ "learning_rate": 9.181279037403387e-05,
2392
+ "loss": 0.0567,
2393
+ "step": 144000
2394
+ },
2395
+ {
2396
+ "epoch": 9.281264050356478,
2397
+ "grad_norm": 0.4280657470226288,
2398
+ "learning_rate": 9.074228702335839e-05,
2399
+ "loss": 0.0554,
2400
+ "step": 144500
2401
+ },
2402
+ {
2403
+ "epoch": 9.313379150876742,
2404
+ "grad_norm": 0.5005618929862976,
2405
+ "learning_rate": 8.967392467938424e-05,
2406
+ "loss": 0.0552,
2407
+ "step": 145000
2408
+ },
2409
+ {
2410
+ "epoch": 9.345494251397007,
2411
+ "grad_norm": 0.3353951573371887,
2412
+ "learning_rate": 8.860342132870876e-05,
2413
+ "loss": 0.0554,
2414
+ "step": 145500
2415
+ },
2416
+ {
2417
+ "epoch": 9.377609351917272,
2418
+ "grad_norm": 0.4684765934944153,
2419
+ "learning_rate": 8.753291797803328e-05,
2420
+ "loss": 0.055,
2421
+ "step": 146000
2422
+ },
2423
+ {
2424
+ "epoch": 9.40060376388978,
2425
+ "eval_loss": 0.08320420235395432,
2426
+ "eval_runtime": 7.3558,
2427
+ "eval_samples_per_second": 67.973,
2428
+ "eval_steps_per_second": 8.565,
2429
+ "step": 146358
2430
+ },
2431
+ {
2432
+ "epoch": 9.409724452437537,
2433
+ "grad_norm": 0.42264777421951294,
2434
+ "learning_rate": 8.646241462735778e-05,
2435
+ "loss": 0.057,
2436
+ "step": 146500
2437
+ },
2438
+ {
2439
+ "epoch": 9.4418395529578,
2440
+ "grad_norm": 0.5612725615501404,
2441
+ "learning_rate": 8.53919112766823e-05,
2442
+ "loss": 0.0552,
2443
+ "step": 147000
2444
+ },
2445
+ {
2446
+ "epoch": 9.473954653478065,
2447
+ "grad_norm": 0.5042113661766052,
2448
+ "learning_rate": 8.432140792600681e-05,
2449
+ "loss": 0.0554,
2450
+ "step": 147500
2451
+ },
2452
+ {
2453
+ "epoch": 9.50606975399833,
2454
+ "grad_norm": 0.6048631072044373,
2455
+ "learning_rate": 8.325090457533132e-05,
2456
+ "loss": 0.0564,
2457
+ "step": 148000
2458
+ },
2459
+ {
2460
+ "epoch": 9.538184854518594,
2461
+ "grad_norm": 0.4988856017589569,
2462
+ "learning_rate": 8.218040122465584e-05,
2463
+ "loss": 0.0569,
2464
+ "step": 148500
2465
+ },
2466
+ {
2467
+ "epoch": 9.570299955038859,
2468
+ "grad_norm": 0.7412840127944946,
2469
+ "learning_rate": 8.110989787398035e-05,
2470
+ "loss": 0.055,
2471
+ "step": 149000
2472
+ },
2473
+ {
2474
+ "epoch": 9.60061660992999,
2475
+ "eval_loss": 0.08181928098201752,
2476
+ "eval_runtime": 7.3732,
2477
+ "eval_samples_per_second": 67.813,
2478
+ "eval_steps_per_second": 8.544,
2479
+ "step": 149472
2480
+ },
2481
+ {
2482
+ "epoch": 9.602415055559124,
2483
+ "grad_norm": 0.3600381314754486,
2484
+ "learning_rate": 8.003939452330486e-05,
2485
+ "loss": 0.0547,
2486
+ "step": 149500
2487
+ },
2488
+ {
2489
+ "epoch": 9.63453015607939,
2490
+ "grad_norm": 0.7391656637191772,
2491
+ "learning_rate": 7.896889117262938e-05,
2492
+ "loss": 0.0554,
2493
+ "step": 150000
2494
+ },
2495
+ {
2496
+ "epoch": 9.666645256599653,
2497
+ "grad_norm": 0.3824392259120941,
2498
+ "learning_rate": 7.789838782195389e-05,
2499
+ "loss": 0.0566,
2500
+ "step": 150500
2501
+ },
2502
+ {
2503
+ "epoch": 9.698760357119918,
2504
+ "grad_norm": 0.5715683698654175,
2505
+ "learning_rate": 7.68278844712784e-05,
2506
+ "loss": 0.0561,
2507
+ "step": 151000
2508
+ },
2509
+ {
2510
+ "epoch": 9.730875457640183,
2511
+ "grad_norm": 1.135609745979309,
2512
+ "learning_rate": 7.57573811206029e-05,
2513
+ "loss": 0.0555,
2514
+ "step": 151500
2515
+ },
2516
+ {
2517
+ "epoch": 9.762990558160446,
2518
+ "grad_norm": 0.5059698820114136,
2519
+ "learning_rate": 7.468901877662877e-05,
2520
+ "loss": 0.0563,
2521
+ "step": 152000
2522
+ },
2523
+ {
2524
+ "epoch": 9.795105658680711,
2525
+ "grad_norm": 0.3883470594882965,
2526
+ "learning_rate": 7.361851542595329e-05,
2527
+ "loss": 0.0544,
2528
+ "step": 152500
2529
+ },
2530
+ {
2531
+ "epoch": 9.800629455970197,
2532
+ "eval_loss": 0.08078183978796005,
2533
+ "eval_runtime": 7.3679,
2534
+ "eval_samples_per_second": 67.862,
2535
+ "eval_steps_per_second": 8.551,
2536
+ "step": 152586
2537
+ },
2538
+ {
2539
+ "epoch": 9.827220759200976,
2540
+ "grad_norm": 0.6080403923988342,
2541
+ "learning_rate": 7.255015308197916e-05,
2542
+ "loss": 0.0553,
2543
+ "step": 153000
2544
+ },
2545
+ {
2546
+ "epoch": 9.859335859721241,
2547
+ "grad_norm": 0.6893211007118225,
2548
+ "learning_rate": 7.147964973130367e-05,
2549
+ "loss": 0.0547,
2550
+ "step": 153500
2551
+ },
2552
+ {
2553
+ "epoch": 9.891450960241505,
2554
+ "grad_norm": 0.5046783089637756,
2555
+ "learning_rate": 7.040914638062818e-05,
2556
+ "loss": 0.0556,
2557
+ "step": 154000
2558
+ },
2559
+ {
2560
+ "epoch": 9.92356606076177,
2561
+ "grad_norm": 0.4692513644695282,
2562
+ "learning_rate": 6.933864302995268e-05,
2563
+ "loss": 0.0548,
2564
+ "step": 154500
2565
+ },
2566
+ {
2567
+ "epoch": 9.955681161282035,
2568
+ "grad_norm": 0.4648891091346741,
2569
+ "learning_rate": 6.827028068597855e-05,
2570
+ "loss": 0.0553,
2571
+ "step": 155000
2572
+ },
2573
+ {
2574
+ "epoch": 9.9877962618023,
2575
+ "grad_norm": 0.34008997678756714,
2576
+ "learning_rate": 6.719977733530306e-05,
2577
+ "loss": 0.0543,
2578
+ "step": 155500
2579
+ },
2580
+ {
2581
+ "epoch": 10.000642302010405,
2582
+ "eval_loss": 0.07938851416110992,
2583
+ "eval_runtime": 7.4753,
2584
+ "eval_samples_per_second": 66.887,
2585
+ "eval_steps_per_second": 8.428,
2586
+ "step": 155700
2587
+ },
2588
+ {
2589
+ "epoch": 10.019911362322564,
2590
+ "grad_norm": 0.4940495491027832,
2591
+ "learning_rate": 6.612927398462756e-05,
2592
+ "loss": 0.0504,
2593
+ "step": 156000
2594
+ },
2595
+ {
2596
+ "epoch": 10.052026462842829,
2597
+ "grad_norm": 0.3671553134918213,
2598
+ "learning_rate": 6.505877063395208e-05,
2599
+ "loss": 0.0471,
2600
+ "step": 156500
2601
+ },
2602
+ {
2603
+ "epoch": 10.084141563363094,
2604
+ "grad_norm": 0.37860697507858276,
2605
+ "learning_rate": 6.39882672832766e-05,
2606
+ "loss": 0.0465,
2607
+ "step": 157000
2608
+ },
2609
+ {
2610
+ "epoch": 10.116256663883357,
2611
+ "grad_norm": 0.7990055084228516,
2612
+ "learning_rate": 6.291776393260112e-05,
2613
+ "loss": 0.0472,
2614
+ "step": 157500
2615
+ },
2616
+ {
2617
+ "epoch": 10.148371764403622,
2618
+ "grad_norm": 0.4885164499282837,
2619
+ "learning_rate": 6.184726058192563e-05,
2620
+ "loss": 0.0482,
2621
+ "step": 158000
2622
+ },
2623
+ {
2624
+ "epoch": 10.180486864923887,
2625
+ "grad_norm": 0.34183698892593384,
2626
+ "learning_rate": 6.077675723125014e-05,
2627
+ "loss": 0.0473,
2628
+ "step": 158500
2629
+ },
2630
+ {
2631
+ "epoch": 10.200655148050613,
2632
+ "eval_loss": 0.08175881206989288,
2633
+ "eval_runtime": 7.3694,
2634
+ "eval_samples_per_second": 67.848,
2635
+ "eval_steps_per_second": 8.549,
2636
+ "step": 158814
2637
+ },
2638
+ {
2639
+ "epoch": 10.212601965444152,
2640
+ "grad_norm": 0.499287486076355,
2641
+ "learning_rate": 5.9706253880574645e-05,
2642
+ "loss": 0.0472,
2643
+ "step": 159000
2644
+ },
2645
+ {
2646
+ "epoch": 10.244717065964416,
2647
+ "grad_norm": 0.6057455539703369,
2648
+ "learning_rate": 5.863575052989916e-05,
2649
+ "loss": 0.0481,
2650
+ "step": 159500
2651
+ },
2652
+ {
2653
+ "epoch": 10.276832166484681,
2654
+ "grad_norm": 0.3568952679634094,
2655
+ "learning_rate": 5.756738818592502e-05,
2656
+ "loss": 0.0468,
2657
+ "step": 160000
2658
+ },
2659
+ {
2660
+ "epoch": 10.308947267004946,
2661
+ "grad_norm": 0.6317808628082275,
2662
+ "learning_rate": 5.649902584195089e-05,
2663
+ "loss": 0.0471,
2664
+ "step": 160500
2665
+ },
2666
+ {
2667
+ "epoch": 10.341062367525211,
2668
+ "grad_norm": 0.365140825510025,
2669
+ "learning_rate": 5.54285224912754e-05,
2670
+ "loss": 0.0487,
2671
+ "step": 161000
2672
+ },
2673
+ {
2674
+ "epoch": 10.373177468045474,
2675
+ "grad_norm": 0.4990079402923584,
2676
+ "learning_rate": 5.4358019140599916e-05,
2677
+ "loss": 0.0475,
2678
+ "step": 161500
2679
+ },
2680
+ {
2681
+ "epoch": 10.40066799409082,
2682
+ "eval_loss": 0.08322105556726456,
2683
+ "eval_runtime": 7.3638,
2684
+ "eval_samples_per_second": 67.9,
2685
+ "eval_steps_per_second": 8.555,
2686
+ "step": 161928
2687
+ },
2688
+ {
2689
+ "epoch": 10.40529256856574,
2690
+ "grad_norm": 0.33129987120628357,
2691
+ "learning_rate": 5.328751578992442e-05,
2692
+ "loss": 0.0468,
2693
+ "step": 162000
2694
+ },
2695
+ {
2696
+ "epoch": 10.437407669086005,
2697
+ "grad_norm": 0.7197828888893127,
2698
+ "learning_rate": 5.2217012439248936e-05,
2699
+ "loss": 0.0484,
2700
+ "step": 162500
2701
+ },
2702
+ {
2703
+ "epoch": 10.469522769606268,
2704
+ "grad_norm": 0.44908684492111206,
2705
+ "learning_rate": 5.11486500952748e-05,
2706
+ "loss": 0.0475,
2707
+ "step": 163000
2708
+ },
2709
+ {
2710
+ "epoch": 10.501637870126533,
2711
+ "grad_norm": 0.6622865200042725,
2712
+ "learning_rate": 5.0078146744599305e-05,
2713
+ "loss": 0.0473,
2714
+ "step": 163500
2715
+ },
2716
+ {
2717
+ "epoch": 10.533752970646798,
2718
+ "grad_norm": 0.5995833873748779,
2719
+ "learning_rate": 4.9007643393923825e-05,
2720
+ "loss": 0.0481,
2721
+ "step": 164000
2722
+ },
2723
+ {
2724
+ "epoch": 10.565868071167063,
2725
+ "grad_norm": 0.6171983480453491,
2726
+ "learning_rate": 4.793714004324834e-05,
2727
+ "loss": 0.0481,
2728
+ "step": 164500
2729
+ },
2730
+ {
2731
+ "epoch": 10.597983171687327,
2732
+ "grad_norm": 0.7169111371040344,
2733
+ "learning_rate": 4.686663669257285e-05,
2734
+ "loss": 0.0481,
2735
+ "step": 165000
2736
+ },
2737
+ {
2738
+ "epoch": 10.60068084013103,
2739
+ "eval_loss": 0.08312381058931351,
2740
+ "eval_runtime": 7.3652,
2741
+ "eval_samples_per_second": 67.887,
2742
+ "eval_steps_per_second": 8.554,
2743
+ "step": 165042
2744
+ },
2745
+ {
2746
+ "epoch": 10.630098272207592,
2747
+ "grad_norm": 0.6282838582992554,
2748
+ "learning_rate": 4.5796133341897365e-05,
2749
+ "loss": 0.0479,
2750
+ "step": 165500
2751
+ },
2752
+ {
2753
+ "epoch": 10.662213372727857,
2754
+ "grad_norm": 0.4022350311279297,
2755
+ "learning_rate": 4.472562999122187e-05,
2756
+ "loss": 0.0476,
2757
+ "step": 166000
2758
+ },
2759
+ {
2760
+ "epoch": 10.694328473248122,
2761
+ "grad_norm": 0.7050232887268066,
2762
+ "learning_rate": 4.3655126640546385e-05,
2763
+ "loss": 0.0482,
2764
+ "step": 166500
2765
+ },
2766
+ {
2767
+ "epoch": 10.726443573768385,
2768
+ "grad_norm": 0.5510446429252625,
2769
+ "learning_rate": 4.2586764296572254e-05,
2770
+ "loss": 0.0478,
2771
+ "step": 167000
2772
+ },
2773
+ {
2774
+ "epoch": 10.75855867428865,
2775
+ "grad_norm": 0.4659407138824463,
2776
+ "learning_rate": 4.151626094589676e-05,
2777
+ "loss": 0.0474,
2778
+ "step": 167500
2779
+ },
2780
+ {
2781
+ "epoch": 10.790673774808916,
2782
+ "grad_norm": 0.5450594425201416,
2783
+ "learning_rate": 4.0445757595221274e-05,
2784
+ "loss": 0.0466,
2785
+ "step": 168000
2786
+ },
2787
+ {
2788
+ "epoch": 10.800693686171238,
2789
+ "eval_loss": 0.08121231198310852,
2790
+ "eval_runtime": 7.3631,
2791
+ "eval_samples_per_second": 67.906,
2792
+ "eval_steps_per_second": 8.556,
2793
+ "step": 168156
2794
+ },
2795
+ {
2796
+ "epoch": 10.822788875329179,
2797
+ "grad_norm": 0.48903122544288635,
2798
+ "learning_rate": 3.937525424454579e-05,
2799
+ "loss": 0.0495,
2800
+ "step": 168500
2801
+ },
2802
+ {
2803
+ "epoch": 10.854903975849444,
2804
+ "grad_norm": 0.6342342495918274,
2805
+ "learning_rate": 3.830689190057165e-05,
2806
+ "loss": 0.0459,
2807
+ "step": 169000
2808
+ },
2809
+ {
2810
+ "epoch": 10.88701907636971,
2811
+ "grad_norm": 0.548468291759491,
2812
+ "learning_rate": 3.723638854989616e-05,
2813
+ "loss": 0.0481,
2814
+ "step": 169500
2815
+ },
2816
+ {
2817
+ "epoch": 10.919134176889974,
2818
+ "grad_norm": 0.776435136795044,
2819
+ "learning_rate": 3.616588519922067e-05,
2820
+ "loss": 0.0476,
2821
+ "step": 170000
2822
+ },
2823
+ {
2824
+ "epoch": 10.951249277410238,
2825
+ "grad_norm": 0.5325737595558167,
2826
+ "learning_rate": 3.509538184854519e-05,
2827
+ "loss": 0.048,
2828
+ "step": 170500
2829
+ },
2830
+ {
2831
+ "epoch": 10.983364377930503,
2832
+ "grad_norm": 0.36368775367736816,
2833
+ "learning_rate": 3.40248784978697e-05,
2834
+ "loss": 0.0461,
2835
+ "step": 171000
2836
+ },
2837
+ {
2838
+ "epoch": 11.000706532211446,
2839
+ "eval_loss": 0.08258914202451706,
2840
+ "eval_runtime": 7.4547,
2841
+ "eval_samples_per_second": 67.071,
2842
+ "eval_steps_per_second": 8.451,
2843
+ "step": 171270
2844
+ },
2845
+ {
2846
+ "epoch": 11.015479478450768,
2847
+ "grad_norm": 0.45326799154281616,
2848
+ "learning_rate": 3.295437514719421e-05,
2849
+ "loss": 0.0454,
2850
+ "step": 171500
2851
+ },
2852
+ {
2853
+ "epoch": 11.047594578971033,
2854
+ "grad_norm": 0.3939746916294098,
2855
+ "learning_rate": 3.188601280322008e-05,
2856
+ "loss": 0.0402,
2857
+ "step": 172000
2858
+ },
2859
+ {
2860
+ "epoch": 11.079709679491296,
2861
+ "grad_norm": 0.3946283459663391,
2862
+ "learning_rate": 3.081765045924594e-05,
2863
+ "loss": 0.0416,
2864
+ "step": 172500
2865
+ },
2866
+ {
2867
+ "epoch": 11.111824780011561,
2868
+ "grad_norm": 0.6206288933753967,
2869
+ "learning_rate": 2.974714710857045e-05,
2870
+ "loss": 0.0405,
2871
+ "step": 173000
2872
+ },
2873
+ {
2874
+ "epoch": 11.143939880531827,
2875
+ "grad_norm": 0.31636786460876465,
2876
+ "learning_rate": 2.867664375789496e-05,
2877
+ "loss": 0.0416,
2878
+ "step": 173500
2879
+ },
2880
+ {
2881
+ "epoch": 11.17605498105209,
2882
+ "grad_norm": 0.5269296765327454,
2883
+ "learning_rate": 2.7606140407219478e-05,
2884
+ "loss": 0.0414,
2885
+ "step": 174000
2886
+ },
2887
+ {
2888
+ "epoch": 11.200719378251653,
2889
+ "eval_loss": 0.0856475681066513,
2890
+ "eval_runtime": 7.3682,
2891
+ "eval_samples_per_second": 67.859,
2892
+ "eval_steps_per_second": 8.55,
2893
+ "step": 174384
2894
+ },
2895
+ {
2896
+ "epoch": 11.208170081572355,
2897
+ "grad_norm": 0.3854086995124817,
2898
+ "learning_rate": 2.6535637056543988e-05,
2899
+ "loss": 0.0402,
2900
+ "step": 174500
2901
+ },
2902
+ {
2903
+ "epoch": 11.24028518209262,
2904
+ "grad_norm": 0.44742342829704285,
2905
+ "learning_rate": 2.5465133705868498e-05,
2906
+ "loss": 0.0415,
2907
+ "step": 175000
2908
+ },
2909
+ {
2910
+ "epoch": 11.272400282612885,
2911
+ "grad_norm": 0.5908629894256592,
2912
+ "learning_rate": 2.439463035519301e-05,
2913
+ "loss": 0.0412,
2914
+ "step": 175500
2915
+ },
2916
+ {
2917
+ "epoch": 11.304515383133149,
2918
+ "grad_norm": 0.562322199344635,
2919
+ "learning_rate": 2.3324127004517528e-05,
2920
+ "loss": 0.0422,
2921
+ "step": 176000
2922
+ },
2923
+ {
2924
+ "epoch": 11.336630483653414,
2925
+ "grad_norm": 0.4340158700942993,
2926
+ "learning_rate": 2.2253623653842038e-05,
2927
+ "loss": 0.041,
2928
+ "step": 176500
2929
+ },
2930
+ {
2931
+ "epoch": 11.368745584173679,
2932
+ "grad_norm": 0.4807555377483368,
2933
+ "learning_rate": 2.118312030316655e-05,
2934
+ "loss": 0.0417,
2935
+ "step": 177000
2936
+ },
2937
+ {
2938
+ "epoch": 11.400732224291861,
2939
+ "eval_loss": 0.08503352105617523,
2940
+ "eval_runtime": 7.3718,
2941
+ "eval_samples_per_second": 67.826,
2942
+ "eval_steps_per_second": 8.546,
2943
+ "step": 177498
2944
+ },
2945
+ {
2946
+ "epoch": 11.400860684693942,
2947
+ "grad_norm": 0.5869391560554504,
2948
+ "learning_rate": 2.011261695249106e-05,
2949
+ "loss": 0.0409,
2950
+ "step": 177500
2951
+ },
2952
+ {
2953
+ "epoch": 11.432975785214207,
2954
+ "grad_norm": 0.323085218667984,
2955
+ "learning_rate": 1.9042113601815575e-05,
2956
+ "loss": 0.0407,
2957
+ "step": 178000
2958
+ },
2959
+ {
2960
+ "epoch": 11.465090885734472,
2961
+ "grad_norm": 0.27529117465019226,
2962
+ "learning_rate": 1.7971610251140085e-05,
2963
+ "loss": 0.0405,
2964
+ "step": 178500
2965
+ },
2966
+ {
2967
+ "epoch": 11.497205986254738,
2968
+ "grad_norm": 0.5203765034675598,
2969
+ "learning_rate": 1.69011069004646e-05,
2970
+ "loss": 0.0411,
2971
+ "step": 179000
2972
+ },
2973
+ {
2974
+ "epoch": 11.529321086775001,
2975
+ "grad_norm": 0.32801443338394165,
2976
+ "learning_rate": 1.5830603549789112e-05,
2977
+ "loss": 0.0406,
2978
+ "step": 179500
2979
+ },
2980
+ {
2981
+ "epoch": 11.561436187295266,
2982
+ "grad_norm": 0.7384726405143738,
2983
+ "learning_rate": 1.4760100199113625e-05,
2984
+ "loss": 0.0414,
2985
+ "step": 180000
2986
+ },
2987
+ {
2988
+ "epoch": 11.593551287815531,
2989
+ "grad_norm": 0.3146156966686249,
2990
+ "learning_rate": 1.3691737855139485e-05,
2991
+ "loss": 0.0404,
2992
+ "step": 180500
2993
+ },
2994
+ {
2995
+ "epoch": 11.60074507033207,
2996
+ "eval_loss": 0.08315666764974594,
2997
+ "eval_runtime": 7.3637,
2998
+ "eval_samples_per_second": 67.9,
2999
+ "eval_steps_per_second": 8.555,
3000
+ "step": 180612
3001
+ },
3002
+ {
3003
+ "epoch": 11.625666388335796,
3004
+ "grad_norm": 0.6330950260162354,
3005
+ "learning_rate": 1.2621234504463999e-05,
3006
+ "loss": 0.0408,
3007
+ "step": 181000
3008
+ },
3009
+ {
3010
+ "epoch": 11.65778148885606,
3011
+ "grad_norm": 0.5293929576873779,
3012
+ "learning_rate": 1.1550731153788512e-05,
3013
+ "loss": 0.04,
3014
+ "step": 181500
3015
+ },
3016
+ {
3017
+ "epoch": 11.689896589376325,
3018
+ "grad_norm": 0.4793175160884857,
3019
+ "learning_rate": 1.0480227803113024e-05,
3020
+ "loss": 0.0399,
3021
+ "step": 182000
3022
+ },
3023
+ {
3024
+ "epoch": 11.72201168989659,
3025
+ "grad_norm": 0.49973323941230774,
3026
+ "learning_rate": 9.409724452437536e-06,
3027
+ "loss": 0.0406,
3028
+ "step": 182500
3029
+ },
3030
+ {
3031
+ "epoch": 11.754126790416855,
3032
+ "grad_norm": 0.4228206276893616,
3033
+ "learning_rate": 8.341362108463401e-06,
3034
+ "loss": 0.0408,
3035
+ "step": 183000
3036
+ },
3037
+ {
3038
+ "epoch": 11.786241890937118,
3039
+ "grad_norm": 0.6146075129508972,
3040
+ "learning_rate": 7.270858757787913e-06,
3041
+ "loss": 0.0413,
3042
+ "step": 183500
3043
+ },
3044
+ {
3045
+ "epoch": 11.800757916372278,
3046
+ "eval_loss": 0.08517524600028992,
3047
+ "eval_runtime": 7.357,
3048
+ "eval_samples_per_second": 67.963,
3049
+ "eval_steps_per_second": 8.563,
3050
+ "step": 183726
3051
+ },
3052
+ {
3053
+ "epoch": 11.818356991457383,
3054
+ "grad_norm": 0.24884289503097534,
3055
+ "learning_rate": 6.200355407112425e-06,
3056
+ "loss": 0.041,
3057
+ "step": 184000
3058
+ },
3059
+ {
3060
+ "epoch": 11.850472091977649,
3061
+ "grad_norm": 0.5679720640182495,
3062
+ "learning_rate": 5.129852056436936e-06,
3063
+ "loss": 0.041,
3064
+ "step": 184500
3065
+ },
3066
+ {
3067
+ "epoch": 11.882587192497912,
3068
+ "grad_norm": 0.553493082523346,
3069
+ "learning_rate": 4.063630719164151e-06,
3070
+ "loss": 0.0398,
3071
+ "step": 185000
3072
+ },
3073
+ {
3074
+ "epoch": 11.914702293018177,
3075
+ "grad_norm": 0.598904550075531,
3076
+ "learning_rate": 2.9931273684886634e-06,
3077
+ "loss": 0.0403,
3078
+ "step": 185500
3079
+ },
3080
+ {
3081
+ "epoch": 11.946817393538442,
3082
+ "grad_norm": 0.6984330415725708,
3083
+ "learning_rate": 1.922624017813176e-06,
3084
+ "loss": 0.0403,
3085
+ "step": 186000
3086
+ },
3087
+ {
3088
+ "epoch": 11.978932494058707,
3089
+ "grad_norm": 0.3562464416027069,
3090
+ "learning_rate": 8.521206671376881e-07,
3091
+ "loss": 0.0411,
3092
+ "step": 186500
3093
+ }
3094
+ ],
3095
+ "logging_steps": 500,
3096
+ "max_steps": 186828,
3097
+ "num_input_tokens_seen": 0,
3098
+ "num_train_epochs": 12,
3099
+ "save_steps": 500,
3100
+ "stateful_callbacks": {
3101
+ "TrainerControl": {
3102
+ "args": {
3103
+ "should_epoch_stop": false,
3104
+ "should_evaluate": false,
3105
+ "should_log": false,
3106
+ "should_save": true,
3107
+ "should_training_stop": true
3108
+ },
3109
+ "attributes": {}
3110
+ }
3111
+ },
3112
+ "total_flos": 4.377469856626455e+19,
3113
+ "train_batch_size": 4,
3114
+ "trial_name": null,
3115
+ "trial_params": null
3116
+ }
checkpoint-186828/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:55428e8dd30daaff792c433246961305cfd20b6b3d93eb34b1e2af5f631a34db
3
+ size 5496