Hugging Face's logo

vidore
/
bipali-hard-v1.1

Safetensors
English
vidore
Model card Files Community
manu commited on
Commit
b6dc9c6
•
1 Parent(s): d214d56

Upload folder using huggingface_hub

Browse files
Files changed (19) hide show
  1. .gitattributes +1 -0
  2. README.md +201 -0
  3. adapter_config.json +26 -0
  4. adapter_model.safetensors +3 -0
  5. checkpoint-3500/README.md +202 -0
  6. checkpoint-3500/adapter_config.json +26 -0
  7. checkpoint-3500/adapter_model.safetensors +3 -0
  8. checkpoint-3500/optimizer.pt +3 -0
  9. checkpoint-3500/rng_state.pth +3 -0
  10. checkpoint-3500/scheduler.pt +3 -0
  11. checkpoint-3500/trainer_state.json +3043 -0
  12. checkpoint-3500/training_args.bin +3 -0
  13. git_hash.txt +1 -0
  14. preprocessor_config.json +40 -0
  15. results.json +1 -0
  16. special_tokens_map.json +39 -0
  17. tokenizer.json +3 -0
  18. tokenizer_config.json +0 -0
  19. training_config.yml +41 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,201 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+
5
+ # Model Card for Model ID
6
+
7
+ <!-- Provide a quick summary of what the model is/does. -->
8
+
9
+
10
+
11
+ ## Model Details
12
+
13
+ ### Model Description
14
+
15
+ <!-- Provide a longer summary of what this model is. -->
16
+
17
+
18
+
19
+ - **Developed by:** [More Information Needed]
20
+ - **Funded by [optional]:** [More Information Needed]
21
+ - **Shared by [optional]:** [More Information Needed]
22
+ - **Model type:** [More Information Needed]
23
+ - **Language(s) (NLP):** [More Information Needed]
24
+ - **License:** [More Information Needed]
25
+ - **Finetuned from model [optional]:** [More Information Needed]
26
+
27
+ ### Model Sources [optional]
28
+
29
+ <!-- Provide the basic links for the model. -->
30
+
31
+ - **Repository:** [More Information Needed]
32
+ - **Paper [optional]:** [More Information Needed]
33
+ - **Demo [optional]:** [More Information Needed]
34
+
35
+ ## Uses
36
+
37
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
38
+
39
+ ### Direct Use
40
+
41
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
42
+
43
+ [More Information Needed]
44
+
45
+ ### Downstream Use [optional]
46
+
47
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
48
+
49
+ [More Information Needed]
50
+
51
+ ### Out-of-Scope Use
52
+
53
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
54
+
55
+ [More Information Needed]
56
+
57
+ ## Bias, Risks, and Limitations
58
+
59
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
60
+
61
+ [More Information Needed]
62
+
63
+ ### Recommendations
64
+
65
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
66
+
67
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
68
+
69
+ ## How to Get Started with the Model
70
+
71
+ Use the code below to get started with the model.
72
+
73
+ [More Information Needed]
74
+
75
+ ## Training Details
76
+
77
+ ### Training Data
78
+
79
+ <!-- This should link to a 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. -->
80
+
81
+ [More Information Needed]
82
+
83
+ ### Training Procedure
84
+
85
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
86
+
87
+ #### Preprocessing [optional]
88
+
89
+ [More Information Needed]
90
+
91
+
92
+ #### Training Hyperparameters
93
+
94
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
95
+
96
+ #### Speeds, Sizes, Times [optional]
97
+
98
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
99
+
100
+ [More Information Needed]
101
+
102
+ ## Evaluation
103
+
104
+ <!-- This section describes the evaluation protocols and provides the results. -->
105
+
106
+ ### Testing Data, Factors & Metrics
107
+
108
+ #### Testing Data
109
+
110
+ <!-- This should link to a Dataset Card if possible. -->
111
+
112
+ [More Information Needed]
113
+
114
+ #### Factors
115
+
116
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
117
+
118
+ [More Information Needed]
119
+
120
+ #### Metrics
121
+
122
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
123
+
124
+ [More Information Needed]
125
+
126
+ ### Results
127
+
128
+ [More Information Needed]
129
+
130
+ #### Summary
131
+
132
+
133
+
134
+ ## Model Examination [optional]
135
+
136
+ <!-- Relevant interpretability work for the model goes here -->
137
+
138
+ [More Information Needed]
139
+
140
+ ## Environmental Impact
141
+
142
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
143
+
144
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
145
+
146
+ - **Hardware Type:** [More Information Needed]
147
+ - **Hours used:** [More Information Needed]
148
+ - **Cloud Provider:** [More Information Needed]
149
+ - **Compute Region:** [More Information Needed]
150
+ - **Carbon Emitted:** [More Information Needed]
151
+
152
+ ## Technical Specifications [optional]
153
+
154
+ ### Model Architecture and Objective
155
+
156
+ [More Information Needed]
157
+
158
+ ### Compute Infrastructure
159
+
160
+ [More Information Needed]
161
+
162
+ #### Hardware
163
+
164
+ [More Information Needed]
165
+
166
+ #### Software
167
+
168
+ [More Information Needed]
169
+
170
+ ## Citation [optional]
171
+
172
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
173
+
174
+ **BibTeX:**
175
+
176
+ [More Information Needed]
177
+
178
+ **APA:**
179
+
180
+ [More Information Needed]
181
+
182
+ ## Glossary [optional]
183
+
184
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
185
+
186
+ [More Information Needed]
187
+
188
+ ## More Information [optional]
189
+
190
+ [More Information Needed]
191
+
192
+ ## Model Card Authors [optional]
193
+
194
+ [More Information Needed]
195
+
196
+ ## Model Card Contact
197
+
198
+ [More Information Needed]
199
+ ### Framework versions
200
+
201
+ - PEFT 0.11.1
adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "./models/paligemma-3b-mix-448",
5
+ "bias": "none",
6
+ "fan_in_fan_out": false,
7
+ "inference_mode": true,
8
+ "init_lora_weights": "gaussian",
9
+ "layer_replication": null,
10
+ "layers_pattern": null,
11
+ "layers_to_transform": null,
12
+ "loftq_config": {},
13
+ "lora_alpha": 32,
14
+ "lora_dropout": 0.1,
15
+ "megatron_config": null,
16
+ "megatron_core": "megatron.core",
17
+ "modules_to_save": null,
18
+ "peft_type": "LORA",
19
+ "r": 32,
20
+ "rank_pattern": {},
21
+ "revision": null,
22
+ "target_modules": "(.*(language_model).*(down_proj|gate_proj|up_proj|k_proj|q_proj|v_proj|o_proj).*$)",
23
+ "task_type": "FEATURE_EXTRACTION",
24
+ "use_dora": false,
25
+ "use_rslora": false
26
+ }
adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b5d3e7367fbea22b45f4bf27fbf20ce8675b79f8ad8eb7512092c5b438b1ede2
3
+ size 78485616
checkpoint-3500/README.md ADDED
@@ -0,0 +1,202 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: ./models/paligemma-3b-mix-448
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 0.11.1
checkpoint-3500/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "./models/paligemma-3b-mix-448",
5
+ "bias": "none",
6
+ "fan_in_fan_out": false,
7
+ "inference_mode": true,
8
+ "init_lora_weights": "gaussian",
9
+ "layer_replication": null,
10
+ "layers_pattern": null,
11
+ "layers_to_transform": null,
12
+ "loftq_config": {},
13
+ "lora_alpha": 32,
14
+ "lora_dropout": 0.1,
15
+ "megatron_config": null,
16
+ "megatron_core": "megatron.core",
17
+ "modules_to_save": null,
18
+ "peft_type": "LORA",
19
+ "r": 32,
20
+ "rank_pattern": {},
21
+ "revision": null,
22
+ "target_modules": "(.*(language_model).*(down_proj|gate_proj|up_proj|k_proj|q_proj|v_proj|o_proj).*$)",
23
+ "task_type": "FEATURE_EXTRACTION",
24
+ "use_dora": false,
25
+ "use_rslora": false
26
+ }
checkpoint-3500/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:82a19d82c629fbf347ba78c97926de17f8495dc1ae66f849e632d41a44e18ef8
3
+ size 78485616
checkpoint-3500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:041b5fa7d34029fcbc078a99e13ce4f702ebab929a3e583f62c15779fbfd2ade
3
+ size 157105466
checkpoint-3500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:95adfcc82f0d0385863355184448cf21447ac25cd25add14f26199dec8c1fc08
3
+ size 14244
checkpoint-3500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:81131d95f053d9e7fed5bbc5ad1529f766bf1139862a226f3877ea7587ceb4cd
3
+ size 1064
checkpoint-3500/trainer_state.json ADDED
@@ -0,0 +1,3043 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9516041326808048,
5
+ "eval_steps": 50,
6
+ "global_step": 3500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0027188689505165853,
13
+ "grad_norm": 0.00921630859375,
14
+ "learning_rate": 5e-06,
15
+ "loss": 0.6937,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.005437737901033171,
20
+ "grad_norm": 0.007415771484375,
21
+ "learning_rate": 1e-05,
22
+ "loss": 0.6937,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.008156606851549755,
27
+ "grad_norm": 0.004119873046875,
28
+ "learning_rate": 1.5e-05,
29
+ "loss": 0.6926,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.010875475802066341,
34
+ "grad_norm": 0.00494384765625,
35
+ "learning_rate": 2e-05,
36
+ "loss": 0.6914,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.013594344752582926,
41
+ "grad_norm": 0.005889892578125,
42
+ "learning_rate": 2.5e-05,
43
+ "loss": 0.6914,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.013594344752582926,
48
+ "eval_loss": 0.69140625,
49
+ "eval_runtime": 140.9452,
50
+ "eval_samples_per_second": 3.547,
51
+ "eval_steps_per_second": 0.114,
52
+ "step": 50
53
+ },
54
+ {
55
+ "epoch": 0.01631321370309951,
56
+ "grad_norm": 0.004730224609375,
57
+ "learning_rate": 3e-05,
58
+ "loss": 0.6914,
59
+ "step": 60
60
+ },
61
+ {
62
+ "epoch": 0.019032082653616094,
63
+ "grad_norm": 0.005279541015625,
64
+ "learning_rate": 3.5e-05,
65
+ "loss": 0.6914,
66
+ "step": 70
67
+ },
68
+ {
69
+ "epoch": 0.021750951604132682,
70
+ "grad_norm": 0.0091552734375,
71
+ "learning_rate": 4e-05,
72
+ "loss": 0.6914,
73
+ "step": 80
74
+ },
75
+ {
76
+ "epoch": 0.024469820554649267,
77
+ "grad_norm": 0.015869140625,
78
+ "learning_rate": 4.5e-05,
79
+ "loss": 0.6914,
80
+ "step": 90
81
+ },
82
+ {
83
+ "epoch": 0.027188689505165852,
84
+ "grad_norm": 0.04638671875,
85
+ "learning_rate": 5e-05,
86
+ "loss": 0.6898,
87
+ "step": 100
88
+ },
89
+ {
90
+ "epoch": 0.027188689505165852,
91
+ "eval_loss": 0.6852499842643738,
92
+ "eval_runtime": 111.472,
93
+ "eval_samples_per_second": 4.485,
94
+ "eval_steps_per_second": 0.144,
95
+ "step": 100
96
+ },
97
+ {
98
+ "epoch": 0.029907558455682437,
99
+ "grad_norm": 0.1572265625,
100
+ "learning_rate": 4.9860257126886535e-05,
101
+ "loss": 0.6805,
102
+ "step": 110
103
+ },
104
+ {
105
+ "epoch": 0.03262642740619902,
106
+ "grad_norm": 0.193359375,
107
+ "learning_rate": 4.972051425377306e-05,
108
+ "loss": 0.6742,
109
+ "step": 120
110
+ },
111
+ {
112
+ "epoch": 0.03534529635671561,
113
+ "grad_norm": 0.2890625,
114
+ "learning_rate": 4.9580771380659594e-05,
115
+ "loss": 0.666,
116
+ "step": 130
117
+ },
118
+ {
119
+ "epoch": 0.03806416530723219,
120
+ "grad_norm": 0.232421875,
121
+ "learning_rate": 4.944102850754612e-05,
122
+ "loss": 0.6625,
123
+ "step": 140
124
+ },
125
+ {
126
+ "epoch": 0.040783034257748776,
127
+ "grad_norm": 0.216796875,
128
+ "learning_rate": 4.9301285634432645e-05,
129
+ "loss": 0.6551,
130
+ "step": 150
131
+ },
132
+ {
133
+ "epoch": 0.040783034257748776,
134
+ "eval_loss": 0.6530625224113464,
135
+ "eval_runtime": 107.5766,
136
+ "eval_samples_per_second": 4.648,
137
+ "eval_steps_per_second": 0.149,
138
+ "step": 150
139
+ },
140
+ {
141
+ "epoch": 0.043501903208265365,
142
+ "grad_norm": 0.2392578125,
143
+ "learning_rate": 4.916154276131917e-05,
144
+ "loss": 0.6562,
145
+ "step": 160
146
+ },
147
+ {
148
+ "epoch": 0.046220772158781946,
149
+ "grad_norm": 0.236328125,
150
+ "learning_rate": 4.9021799888205704e-05,
151
+ "loss": 0.6516,
152
+ "step": 170
153
+ },
154
+ {
155
+ "epoch": 0.048939641109298535,
156
+ "grad_norm": 0.2138671875,
157
+ "learning_rate": 4.888205701509223e-05,
158
+ "loss": 0.6512,
159
+ "step": 180
160
+ },
161
+ {
162
+ "epoch": 0.051658510059815116,
163
+ "grad_norm": 0.2021484375,
164
+ "learning_rate": 4.874231414197876e-05,
165
+ "loss": 0.648,
166
+ "step": 190
167
+ },
168
+ {
169
+ "epoch": 0.054377379010331704,
170
+ "grad_norm": 0.1982421875,
171
+ "learning_rate": 4.860257126886529e-05,
172
+ "loss": 0.6391,
173
+ "step": 200
174
+ },
175
+ {
176
+ "epoch": 0.054377379010331704,
177
+ "eval_loss": 0.6452187299728394,
178
+ "eval_runtime": 107.1804,
179
+ "eval_samples_per_second": 4.665,
180
+ "eval_steps_per_second": 0.149,
181
+ "step": 200
182
+ },
183
+ {
184
+ "epoch": 0.057096247960848286,
185
+ "grad_norm": 0.212890625,
186
+ "learning_rate": 4.846282839575182e-05,
187
+ "loss": 0.6445,
188
+ "step": 210
189
+ },
190
+ {
191
+ "epoch": 0.059815116911364874,
192
+ "grad_norm": 0.24609375,
193
+ "learning_rate": 4.8323085522638347e-05,
194
+ "loss": 0.6477,
195
+ "step": 220
196
+ },
197
+ {
198
+ "epoch": 0.06253398586188146,
199
+ "grad_norm": 0.2021484375,
200
+ "learning_rate": 4.818334264952488e-05,
201
+ "loss": 0.6402,
202
+ "step": 230
203
+ },
204
+ {
205
+ "epoch": 0.06525285481239804,
206
+ "grad_norm": 0.2412109375,
207
+ "learning_rate": 4.8043599776411405e-05,
208
+ "loss": 0.641,
209
+ "step": 240
210
+ },
211
+ {
212
+ "epoch": 0.06797172376291463,
213
+ "grad_norm": 0.2177734375,
214
+ "learning_rate": 4.790385690329793e-05,
215
+ "loss": 0.6355,
216
+ "step": 250
217
+ },
218
+ {
219
+ "epoch": 0.06797172376291463,
220
+ "eval_loss": 0.6340000033378601,
221
+ "eval_runtime": 104.7186,
222
+ "eval_samples_per_second": 4.775,
223
+ "eval_steps_per_second": 0.153,
224
+ "step": 250
225
+ },
226
+ {
227
+ "epoch": 0.07069059271343121,
228
+ "grad_norm": 0.25,
229
+ "learning_rate": 4.7764114030184464e-05,
230
+ "loss": 0.6324,
231
+ "step": 260
232
+ },
233
+ {
234
+ "epoch": 0.0734094616639478,
235
+ "grad_norm": 0.24609375,
236
+ "learning_rate": 4.762437115707099e-05,
237
+ "loss": 0.6309,
238
+ "step": 270
239
+ },
240
+ {
241
+ "epoch": 0.07612833061446438,
242
+ "grad_norm": 0.25,
243
+ "learning_rate": 4.748462828395752e-05,
244
+ "loss": 0.6289,
245
+ "step": 280
246
+ },
247
+ {
248
+ "epoch": 0.07884719956498097,
249
+ "grad_norm": 0.2490234375,
250
+ "learning_rate": 4.734488541084405e-05,
251
+ "loss": 0.6273,
252
+ "step": 290
253
+ },
254
+ {
255
+ "epoch": 0.08156606851549755,
256
+ "grad_norm": 0.24609375,
257
+ "learning_rate": 4.720514253773058e-05,
258
+ "loss": 0.6352,
259
+ "step": 300
260
+ },
261
+ {
262
+ "epoch": 0.08156606851549755,
263
+ "eval_loss": 0.628250002861023,
264
+ "eval_runtime": 102.128,
265
+ "eval_samples_per_second": 4.896,
266
+ "eval_steps_per_second": 0.157,
267
+ "step": 300
268
+ },
269
+ {
270
+ "epoch": 0.08428493746601413,
271
+ "grad_norm": 0.244140625,
272
+ "learning_rate": 4.7065399664617106e-05,
273
+ "loss": 0.6297,
274
+ "step": 310
275
+ },
276
+ {
277
+ "epoch": 0.08700380641653073,
278
+ "grad_norm": 0.251953125,
279
+ "learning_rate": 4.692565679150364e-05,
280
+ "loss": 0.6184,
281
+ "step": 320
282
+ },
283
+ {
284
+ "epoch": 0.08972267536704731,
285
+ "grad_norm": 0.23046875,
286
+ "learning_rate": 4.6785913918390165e-05,
287
+ "loss": 0.623,
288
+ "step": 330
289
+ },
290
+ {
291
+ "epoch": 0.09244154431756389,
292
+ "grad_norm": 0.29296875,
293
+ "learning_rate": 4.664617104527669e-05,
294
+ "loss": 0.6246,
295
+ "step": 340
296
+ },
297
+ {
298
+ "epoch": 0.09516041326808047,
299
+ "grad_norm": 0.2314453125,
300
+ "learning_rate": 4.650642817216322e-05,
301
+ "loss": 0.6293,
302
+ "step": 350
303
+ },
304
+ {
305
+ "epoch": 0.09516041326808047,
306
+ "eval_loss": 0.6250312328338623,
307
+ "eval_runtime": 133.4332,
308
+ "eval_samples_per_second": 3.747,
309
+ "eval_steps_per_second": 0.12,
310
+ "step": 350
311
+ },
312
+ {
313
+ "epoch": 0.09787928221859707,
314
+ "grad_norm": 0.244140625,
315
+ "learning_rate": 4.636668529904975e-05,
316
+ "loss": 0.6324,
317
+ "step": 360
318
+ },
319
+ {
320
+ "epoch": 0.10059815116911365,
321
+ "grad_norm": 0.212890625,
322
+ "learning_rate": 4.622694242593628e-05,
323
+ "loss": 0.6145,
324
+ "step": 370
325
+ },
326
+ {
327
+ "epoch": 0.10331702011963023,
328
+ "grad_norm": 0.2890625,
329
+ "learning_rate": 4.608719955282281e-05,
330
+ "loss": 0.6297,
331
+ "step": 380
332
+ },
333
+ {
334
+ "epoch": 0.10603588907014681,
335
+ "grad_norm": 0.30859375,
336
+ "learning_rate": 4.594745667970934e-05,
337
+ "loss": 0.627,
338
+ "step": 390
339
+ },
340
+ {
341
+ "epoch": 0.10875475802066341,
342
+ "grad_norm": 0.2119140625,
343
+ "learning_rate": 4.5807713806595866e-05,
344
+ "loss": 0.6254,
345
+ "step": 400
346
+ },
347
+ {
348
+ "epoch": 0.10875475802066341,
349
+ "eval_loss": 0.6199687719345093,
350
+ "eval_runtime": 110.9675,
351
+ "eval_samples_per_second": 4.506,
352
+ "eval_steps_per_second": 0.144,
353
+ "step": 400
354
+ },
355
+ {
356
+ "epoch": 0.11147362697117999,
357
+ "grad_norm": 0.22265625,
358
+ "learning_rate": 4.56679709334824e-05,
359
+ "loss": 0.6219,
360
+ "step": 410
361
+ },
362
+ {
363
+ "epoch": 0.11419249592169657,
364
+ "grad_norm": 0.2294921875,
365
+ "learning_rate": 4.5528228060368925e-05,
366
+ "loss": 0.6258,
367
+ "step": 420
368
+ },
369
+ {
370
+ "epoch": 0.11691136487221315,
371
+ "grad_norm": 0.248046875,
372
+ "learning_rate": 4.538848518725545e-05,
373
+ "loss": 0.6266,
374
+ "step": 430
375
+ },
376
+ {
377
+ "epoch": 0.11963023382272975,
378
+ "grad_norm": 0.2109375,
379
+ "learning_rate": 4.5248742314141976e-05,
380
+ "loss": 0.6258,
381
+ "step": 440
382
+ },
383
+ {
384
+ "epoch": 0.12234910277324633,
385
+ "grad_norm": 0.228515625,
386
+ "learning_rate": 4.510899944102851e-05,
387
+ "loss": 0.6238,
388
+ "step": 450
389
+ },
390
+ {
391
+ "epoch": 0.12234910277324633,
392
+ "eval_loss": 0.617062509059906,
393
+ "eval_runtime": 109.0982,
394
+ "eval_samples_per_second": 4.583,
395
+ "eval_steps_per_second": 0.147,
396
+ "step": 450
397
+ },
398
+ {
399
+ "epoch": 0.1250679717237629,
400
+ "grad_norm": 0.2490234375,
401
+ "learning_rate": 4.4969256567915035e-05,
402
+ "loss": 0.616,
403
+ "step": 460
404
+ },
405
+ {
406
+ "epoch": 0.1277868406742795,
407
+ "grad_norm": 0.2197265625,
408
+ "learning_rate": 4.482951369480157e-05,
409
+ "loss": 0.6172,
410
+ "step": 470
411
+ },
412
+ {
413
+ "epoch": 0.13050570962479607,
414
+ "grad_norm": 0.373046875,
415
+ "learning_rate": 4.46897708216881e-05,
416
+ "loss": 0.6199,
417
+ "step": 480
418
+ },
419
+ {
420
+ "epoch": 0.13322457857531267,
421
+ "grad_norm": 0.2470703125,
422
+ "learning_rate": 4.4550027948574626e-05,
423
+ "loss": 0.6172,
424
+ "step": 490
425
+ },
426
+ {
427
+ "epoch": 0.13594344752582926,
428
+ "grad_norm": 0.2265625,
429
+ "learning_rate": 4.441028507546116e-05,
430
+ "loss": 0.6152,
431
+ "step": 500
432
+ },
433
+ {
434
+ "epoch": 0.13594344752582926,
435
+ "eval_loss": 0.6169062256813049,
436
+ "eval_runtime": 106.3936,
437
+ "eval_samples_per_second": 4.7,
438
+ "eval_steps_per_second": 0.15,
439
+ "step": 500
440
+ },
441
+ {
442
+ "epoch": 0.13866231647634583,
443
+ "grad_norm": 0.2353515625,
444
+ "learning_rate": 4.4270542202347684e-05,
445
+ "loss": 0.6125,
446
+ "step": 510
447
+ },
448
+ {
449
+ "epoch": 0.14138118542686243,
450
+ "grad_norm": 0.29296875,
451
+ "learning_rate": 4.413079932923421e-05,
452
+ "loss": 0.6176,
453
+ "step": 520
454
+ },
455
+ {
456
+ "epoch": 0.14410005437737902,
457
+ "grad_norm": 0.263671875,
458
+ "learning_rate": 4.3991056456120736e-05,
459
+ "loss": 0.6207,
460
+ "step": 530
461
+ },
462
+ {
463
+ "epoch": 0.1468189233278956,
464
+ "grad_norm": 0.244140625,
465
+ "learning_rate": 4.385131358300727e-05,
466
+ "loss": 0.6172,
467
+ "step": 540
468
+ },
469
+ {
470
+ "epoch": 0.14953779227841218,
471
+ "grad_norm": 0.2275390625,
472
+ "learning_rate": 4.3711570709893795e-05,
473
+ "loss": 0.6207,
474
+ "step": 550
475
+ },
476
+ {
477
+ "epoch": 0.14953779227841218,
478
+ "eval_loss": 0.6126562356948853,
479
+ "eval_runtime": 104.0346,
480
+ "eval_samples_per_second": 4.806,
481
+ "eval_steps_per_second": 0.154,
482
+ "step": 550
483
+ },
484
+ {
485
+ "epoch": 0.15225666122892875,
486
+ "grad_norm": 0.265625,
487
+ "learning_rate": 4.357182783678033e-05,
488
+ "loss": 0.6094,
489
+ "step": 560
490
+ },
491
+ {
492
+ "epoch": 0.15497553017944535,
493
+ "grad_norm": 0.236328125,
494
+ "learning_rate": 4.343208496366685e-05,
495
+ "loss": 0.6098,
496
+ "step": 570
497
+ },
498
+ {
499
+ "epoch": 0.15769439912996194,
500
+ "grad_norm": 0.24609375,
501
+ "learning_rate": 4.3292342090553386e-05,
502
+ "loss": 0.6156,
503
+ "step": 580
504
+ },
505
+ {
506
+ "epoch": 0.1604132680804785,
507
+ "grad_norm": 0.265625,
508
+ "learning_rate": 4.315259921743992e-05,
509
+ "loss": 0.6191,
510
+ "step": 590
511
+ },
512
+ {
513
+ "epoch": 0.1631321370309951,
514
+ "grad_norm": 0.2333984375,
515
+ "learning_rate": 4.3012856344326444e-05,
516
+ "loss": 0.6129,
517
+ "step": 600
518
+ },
519
+ {
520
+ "epoch": 0.1631321370309951,
521
+ "eval_loss": 0.6111562252044678,
522
+ "eval_runtime": 101.9682,
523
+ "eval_samples_per_second": 4.903,
524
+ "eval_steps_per_second": 0.157,
525
+ "step": 600
526
+ },
527
+ {
528
+ "epoch": 0.1658510059815117,
529
+ "grad_norm": 0.234375,
530
+ "learning_rate": 4.287311347121297e-05,
531
+ "loss": 0.6168,
532
+ "step": 610
533
+ },
534
+ {
535
+ "epoch": 0.16856987493202827,
536
+ "grad_norm": 0.21484375,
537
+ "learning_rate": 4.2733370598099496e-05,
538
+ "loss": 0.609,
539
+ "step": 620
540
+ },
541
+ {
542
+ "epoch": 0.17128874388254486,
543
+ "grad_norm": 0.26953125,
544
+ "learning_rate": 4.259362772498603e-05,
545
+ "loss": 0.6168,
546
+ "step": 630
547
+ },
548
+ {
549
+ "epoch": 0.17400761283306146,
550
+ "grad_norm": 0.232421875,
551
+ "learning_rate": 4.2453884851872554e-05,
552
+ "loss": 0.6125,
553
+ "step": 640
554
+ },
555
+ {
556
+ "epoch": 0.17672648178357803,
557
+ "grad_norm": 0.23828125,
558
+ "learning_rate": 4.231414197875909e-05,
559
+ "loss": 0.6148,
560
+ "step": 650
561
+ },
562
+ {
563
+ "epoch": 0.17672648178357803,
564
+ "eval_loss": 0.6060624718666077,
565
+ "eval_runtime": 101.5093,
566
+ "eval_samples_per_second": 4.926,
567
+ "eval_steps_per_second": 0.158,
568
+ "step": 650
569
+ },
570
+ {
571
+ "epoch": 0.17944535073409462,
572
+ "grad_norm": 0.232421875,
573
+ "learning_rate": 4.217439910564561e-05,
574
+ "loss": 0.6156,
575
+ "step": 660
576
+ },
577
+ {
578
+ "epoch": 0.1821642196846112,
579
+ "grad_norm": 0.23828125,
580
+ "learning_rate": 4.2034656232532146e-05,
581
+ "loss": 0.6172,
582
+ "step": 670
583
+ },
584
+ {
585
+ "epoch": 0.18488308863512778,
586
+ "grad_norm": 0.2255859375,
587
+ "learning_rate": 4.189491335941867e-05,
588
+ "loss": 0.6148,
589
+ "step": 680
590
+ },
591
+ {
592
+ "epoch": 0.18760195758564438,
593
+ "grad_norm": 0.232421875,
594
+ "learning_rate": 4.17551704863052e-05,
595
+ "loss": 0.6148,
596
+ "step": 690
597
+ },
598
+ {
599
+ "epoch": 0.19032082653616095,
600
+ "grad_norm": 0.2392578125,
601
+ "learning_rate": 4.161542761319172e-05,
602
+ "loss": 0.6039,
603
+ "step": 700
604
+ },
605
+ {
606
+ "epoch": 0.19032082653616095,
607
+ "eval_loss": 0.6069062352180481,
608
+ "eval_runtime": 101.1705,
609
+ "eval_samples_per_second": 4.942,
610
+ "eval_steps_per_second": 0.158,
611
+ "step": 700
612
+ },
613
+ {
614
+ "epoch": 0.19303969548667754,
615
+ "grad_norm": 0.244140625,
616
+ "learning_rate": 4.1475684740078256e-05,
617
+ "loss": 0.6086,
618
+ "step": 710
619
+ },
620
+ {
621
+ "epoch": 0.19575856443719414,
622
+ "grad_norm": 0.275390625,
623
+ "learning_rate": 4.133594186696479e-05,
624
+ "loss": 0.6137,
625
+ "step": 720
626
+ },
627
+ {
628
+ "epoch": 0.1984774333877107,
629
+ "grad_norm": 0.259765625,
630
+ "learning_rate": 4.1196198993851314e-05,
631
+ "loss": 0.6168,
632
+ "step": 730
633
+ },
634
+ {
635
+ "epoch": 0.2011963023382273,
636
+ "grad_norm": 0.2451171875,
637
+ "learning_rate": 4.105645612073785e-05,
638
+ "loss": 0.602,
639
+ "step": 740
640
+ },
641
+ {
642
+ "epoch": 0.2039151712887439,
643
+ "grad_norm": 0.263671875,
644
+ "learning_rate": 4.091671324762437e-05,
645
+ "loss": 0.6027,
646
+ "step": 750
647
+ },
648
+ {
649
+ "epoch": 0.2039151712887439,
650
+ "eval_loss": 0.6099687218666077,
651
+ "eval_runtime": 98.7786,
652
+ "eval_samples_per_second": 5.062,
653
+ "eval_steps_per_second": 0.162,
654
+ "step": 750
655
+ },
656
+ {
657
+ "epoch": 0.20663404023926046,
658
+ "grad_norm": 0.2490234375,
659
+ "learning_rate": 4.0776970374510905e-05,
660
+ "loss": 0.609,
661
+ "step": 760
662
+ },
663
+ {
664
+ "epoch": 0.20935290918977706,
665
+ "grad_norm": 0.23046875,
666
+ "learning_rate": 4.063722750139743e-05,
667
+ "loss": 0.6023,
668
+ "step": 770
669
+ },
670
+ {
671
+ "epoch": 0.21207177814029363,
672
+ "grad_norm": 0.283203125,
673
+ "learning_rate": 4.049748462828396e-05,
674
+ "loss": 0.6066,
675
+ "step": 780
676
+ },
677
+ {
678
+ "epoch": 0.21479064709081022,
679
+ "grad_norm": 0.240234375,
680
+ "learning_rate": 4.035774175517048e-05,
681
+ "loss": 0.6109,
682
+ "step": 790
683
+ },
684
+ {
685
+ "epoch": 0.21750951604132682,
686
+ "grad_norm": 0.2421875,
687
+ "learning_rate": 4.0217998882057016e-05,
688
+ "loss": 0.6062,
689
+ "step": 800
690
+ },
691
+ {
692
+ "epoch": 0.21750951604132682,
693
+ "eval_loss": 0.6081562638282776,
694
+ "eval_runtime": 98.0178,
695
+ "eval_samples_per_second": 5.101,
696
+ "eval_steps_per_second": 0.163,
697
+ "step": 800
698
+ },
699
+ {
700
+ "epoch": 0.22022838499184338,
701
+ "grad_norm": 0.2275390625,
702
+ "learning_rate": 4.007825600894354e-05,
703
+ "loss": 0.6109,
704
+ "step": 810
705
+ },
706
+ {
707
+ "epoch": 0.22294725394235998,
708
+ "grad_norm": 0.2421875,
709
+ "learning_rate": 3.9938513135830074e-05,
710
+ "loss": 0.6082,
711
+ "step": 820
712
+ },
713
+ {
714
+ "epoch": 0.22566612289287658,
715
+ "grad_norm": 0.2451171875,
716
+ "learning_rate": 3.979877026271661e-05,
717
+ "loss": 0.6059,
718
+ "step": 830
719
+ },
720
+ {
721
+ "epoch": 0.22838499184339314,
722
+ "grad_norm": 0.240234375,
723
+ "learning_rate": 3.965902738960313e-05,
724
+ "loss": 0.6047,
725
+ "step": 840
726
+ },
727
+ {
728
+ "epoch": 0.23110386079390974,
729
+ "grad_norm": 0.2275390625,
730
+ "learning_rate": 3.9519284516489665e-05,
731
+ "loss": 0.6055,
732
+ "step": 850
733
+ },
734
+ {
735
+ "epoch": 0.23110386079390974,
736
+ "eval_loss": 0.6029062271118164,
737
+ "eval_runtime": 97.6493,
738
+ "eval_samples_per_second": 5.12,
739
+ "eval_steps_per_second": 0.164,
740
+ "step": 850
741
+ },
742
+ {
743
+ "epoch": 0.2338227297444263,
744
+ "grad_norm": 0.251953125,
745
+ "learning_rate": 3.937954164337619e-05,
746
+ "loss": 0.6031,
747
+ "step": 860
748
+ },
749
+ {
750
+ "epoch": 0.2365415986949429,
751
+ "grad_norm": 0.294921875,
752
+ "learning_rate": 3.923979877026272e-05,
753
+ "loss": 0.609,
754
+ "step": 870
755
+ },
756
+ {
757
+ "epoch": 0.2392604676454595,
758
+ "grad_norm": 0.240234375,
759
+ "learning_rate": 3.910005589714924e-05,
760
+ "loss": 0.6078,
761
+ "step": 880
762
+ },
763
+ {
764
+ "epoch": 0.24197933659597606,
765
+ "grad_norm": 0.2392578125,
766
+ "learning_rate": 3.8960313024035775e-05,
767
+ "loss": 0.607,
768
+ "step": 890
769
+ },
770
+ {
771
+ "epoch": 0.24469820554649266,
772
+ "grad_norm": 0.2578125,
773
+ "learning_rate": 3.88205701509223e-05,
774
+ "loss": 0.6031,
775
+ "step": 900
776
+ },
777
+ {
778
+ "epoch": 0.24469820554649266,
779
+ "eval_loss": 0.6027500033378601,
780
+ "eval_runtime": 105.7702,
781
+ "eval_samples_per_second": 4.727,
782
+ "eval_steps_per_second": 0.151,
783
+ "step": 900
784
+ },
785
+ {
786
+ "epoch": 0.24741707449700925,
787
+ "grad_norm": 0.2216796875,
788
+ "learning_rate": 3.8680827277808834e-05,
789
+ "loss": 0.607,
790
+ "step": 910
791
+ },
792
+ {
793
+ "epoch": 0.2501359434475258,
794
+ "grad_norm": 0.326171875,
795
+ "learning_rate": 3.854108440469536e-05,
796
+ "loss": 0.5957,
797
+ "step": 920
798
+ },
799
+ {
800
+ "epoch": 0.2528548123980424,
801
+ "grad_norm": 0.26171875,
802
+ "learning_rate": 3.840134153158189e-05,
803
+ "loss": 0.6012,
804
+ "step": 930
805
+ },
806
+ {
807
+ "epoch": 0.255573681348559,
808
+ "grad_norm": 0.259765625,
809
+ "learning_rate": 3.826159865846842e-05,
810
+ "loss": 0.5969,
811
+ "step": 940
812
+ },
813
+ {
814
+ "epoch": 0.2582925502990756,
815
+ "grad_norm": 0.255859375,
816
+ "learning_rate": 3.812185578535495e-05,
817
+ "loss": 0.6047,
818
+ "step": 950
819
+ },
820
+ {
821
+ "epoch": 0.2582925502990756,
822
+ "eval_loss": 0.6035937666893005,
823
+ "eval_runtime": 100.7957,
824
+ "eval_samples_per_second": 4.961,
825
+ "eval_steps_per_second": 0.159,
826
+ "step": 950
827
+ },
828
+ {
829
+ "epoch": 0.26101141924959215,
830
+ "grad_norm": 0.263671875,
831
+ "learning_rate": 3.798211291224148e-05,
832
+ "loss": 0.6035,
833
+ "step": 960
834
+ },
835
+ {
836
+ "epoch": 0.26373028820010874,
837
+ "grad_norm": 0.248046875,
838
+ "learning_rate": 3.7842370039128e-05,
839
+ "loss": 0.6035,
840
+ "step": 970
841
+ },
842
+ {
843
+ "epoch": 0.26644915715062534,
844
+ "grad_norm": 0.25390625,
845
+ "learning_rate": 3.7702627166014535e-05,
846
+ "loss": 0.6078,
847
+ "step": 980
848
+ },
849
+ {
850
+ "epoch": 0.26916802610114193,
851
+ "grad_norm": 0.5078125,
852
+ "learning_rate": 3.756288429290106e-05,
853
+ "loss": 0.6031,
854
+ "step": 990
855
+ },
856
+ {
857
+ "epoch": 0.27188689505165853,
858
+ "grad_norm": 0.248046875,
859
+ "learning_rate": 3.7423141419787594e-05,
860
+ "loss": 0.616,
861
+ "step": 1000
862
+ },
863
+ {
864
+ "epoch": 0.27188689505165853,
865
+ "eval_loss": 0.6025000214576721,
866
+ "eval_runtime": 99.8927,
867
+ "eval_samples_per_second": 5.005,
868
+ "eval_steps_per_second": 0.16,
869
+ "step": 1000
870
+ },
871
+ {
872
+ "epoch": 0.27460576400217507,
873
+ "grad_norm": 0.287109375,
874
+ "learning_rate": 3.728339854667412e-05,
875
+ "loss": 0.6016,
876
+ "step": 1010
877
+ },
878
+ {
879
+ "epoch": 0.27732463295269166,
880
+ "grad_norm": 0.2314453125,
881
+ "learning_rate": 3.714365567356065e-05,
882
+ "loss": 0.5992,
883
+ "step": 1020
884
+ },
885
+ {
886
+ "epoch": 0.28004350190320826,
887
+ "grad_norm": 0.2392578125,
888
+ "learning_rate": 3.700391280044718e-05,
889
+ "loss": 0.6047,
890
+ "step": 1030
891
+ },
892
+ {
893
+ "epoch": 0.28276237085372485,
894
+ "grad_norm": 0.263671875,
895
+ "learning_rate": 3.686416992733371e-05,
896
+ "loss": 0.6035,
897
+ "step": 1040
898
+ },
899
+ {
900
+ "epoch": 0.28548123980424145,
901
+ "grad_norm": 0.26171875,
902
+ "learning_rate": 3.6724427054220237e-05,
903
+ "loss": 0.6055,
904
+ "step": 1050
905
+ },
906
+ {
907
+ "epoch": 0.28548123980424145,
908
+ "eval_loss": 0.6012499928474426,
909
+ "eval_runtime": 100.4493,
910
+ "eval_samples_per_second": 4.978,
911
+ "eval_steps_per_second": 0.159,
912
+ "step": 1050
913
+ },
914
+ {
915
+ "epoch": 0.28820010875475804,
916
+ "grad_norm": 0.255859375,
917
+ "learning_rate": 3.658468418110676e-05,
918
+ "loss": 0.6098,
919
+ "step": 1060
920
+ },
921
+ {
922
+ "epoch": 0.2909189777052746,
923
+ "grad_norm": 0.310546875,
924
+ "learning_rate": 3.6444941307993295e-05,
925
+ "loss": 0.5941,
926
+ "step": 1070
927
+ },
928
+ {
929
+ "epoch": 0.2936378466557912,
930
+ "grad_norm": 0.25390625,
931
+ "learning_rate": 3.630519843487982e-05,
932
+ "loss": 0.602,
933
+ "step": 1080
934
+ },
935
+ {
936
+ "epoch": 0.2963567156063078,
937
+ "grad_norm": 0.27734375,
938
+ "learning_rate": 3.6165455561766354e-05,
939
+ "loss": 0.5992,
940
+ "step": 1090
941
+ },
942
+ {
943
+ "epoch": 0.29907558455682437,
944
+ "grad_norm": 0.5859375,
945
+ "learning_rate": 3.602571268865288e-05,
946
+ "loss": 0.609,
947
+ "step": 1100
948
+ },
949
+ {
950
+ "epoch": 0.29907558455682437,
951
+ "eval_loss": 0.5996875166893005,
952
+ "eval_runtime": 100.6816,
953
+ "eval_samples_per_second": 4.966,
954
+ "eval_steps_per_second": 0.159,
955
+ "step": 1100
956
+ },
957
+ {
958
+ "epoch": 0.30179445350734097,
959
+ "grad_norm": 0.259765625,
960
+ "learning_rate": 3.588596981553941e-05,
961
+ "loss": 0.6047,
962
+ "step": 1110
963
+ },
964
+ {
965
+ "epoch": 0.3045133224578575,
966
+ "grad_norm": 0.255859375,
967
+ "learning_rate": 3.574622694242594e-05,
968
+ "loss": 0.6062,
969
+ "step": 1120
970
+ },
971
+ {
972
+ "epoch": 0.3072321914083741,
973
+ "grad_norm": 0.2333984375,
974
+ "learning_rate": 3.560648406931247e-05,
975
+ "loss": 0.6051,
976
+ "step": 1130
977
+ },
978
+ {
979
+ "epoch": 0.3099510603588907,
980
+ "grad_norm": 0.427734375,
981
+ "learning_rate": 3.5466741196198996e-05,
982
+ "loss": 0.6102,
983
+ "step": 1140
984
+ },
985
+ {
986
+ "epoch": 0.3126699293094073,
987
+ "grad_norm": 0.265625,
988
+ "learning_rate": 3.532699832308552e-05,
989
+ "loss": 0.6035,
990
+ "step": 1150
991
+ },
992
+ {
993
+ "epoch": 0.3126699293094073,
994
+ "eval_loss": 0.5978124737739563,
995
+ "eval_runtime": 98.5391,
996
+ "eval_samples_per_second": 5.074,
997
+ "eval_steps_per_second": 0.162,
998
+ "step": 1150
999
+ },
1000
+ {
1001
+ "epoch": 0.3153887982599239,
1002
+ "grad_norm": 0.263671875,
1003
+ "learning_rate": 3.518725544997205e-05,
1004
+ "loss": 0.6023,
1005
+ "step": 1160
1006
+ },
1007
+ {
1008
+ "epoch": 0.3181076672104405,
1009
+ "grad_norm": 0.275390625,
1010
+ "learning_rate": 3.504751257685858e-05,
1011
+ "loss": 0.6082,
1012
+ "step": 1170
1013
+ },
1014
+ {
1015
+ "epoch": 0.320826536160957,
1016
+ "grad_norm": 0.26171875,
1017
+ "learning_rate": 3.4907769703745107e-05,
1018
+ "loss": 0.5973,
1019
+ "step": 1180
1020
+ },
1021
+ {
1022
+ "epoch": 0.3235454051114736,
1023
+ "grad_norm": 0.28515625,
1024
+ "learning_rate": 3.476802683063164e-05,
1025
+ "loss": 0.6027,
1026
+ "step": 1190
1027
+ },
1028
+ {
1029
+ "epoch": 0.3262642740619902,
1030
+ "grad_norm": 0.263671875,
1031
+ "learning_rate": 3.462828395751817e-05,
1032
+ "loss": 0.5984,
1033
+ "step": 1200
1034
+ },
1035
+ {
1036
+ "epoch": 0.3262642740619902,
1037
+ "eval_loss": 0.5985000133514404,
1038
+ "eval_runtime": 99.1902,
1039
+ "eval_samples_per_second": 5.041,
1040
+ "eval_steps_per_second": 0.161,
1041
+ "step": 1200
1042
+ },
1043
+ {
1044
+ "epoch": 0.3289831430125068,
1045
+ "grad_norm": 0.298828125,
1046
+ "learning_rate": 3.44885410844047e-05,
1047
+ "loss": 0.5992,
1048
+ "step": 1210
1049
+ },
1050
+ {
1051
+ "epoch": 0.3317020119630234,
1052
+ "grad_norm": 0.2490234375,
1053
+ "learning_rate": 3.434879821129123e-05,
1054
+ "loss": 0.6125,
1055
+ "step": 1220
1056
+ },
1057
+ {
1058
+ "epoch": 0.33442088091353994,
1059
+ "grad_norm": 0.2490234375,
1060
+ "learning_rate": 3.4209055338177756e-05,
1061
+ "loss": 0.5984,
1062
+ "step": 1230
1063
+ },
1064
+ {
1065
+ "epoch": 0.33713974986405654,
1066
+ "grad_norm": 0.2333984375,
1067
+ "learning_rate": 3.406931246506428e-05,
1068
+ "loss": 0.6,
1069
+ "step": 1240
1070
+ },
1071
+ {
1072
+ "epoch": 0.33985861881457313,
1073
+ "grad_norm": 0.28515625,
1074
+ "learning_rate": 3.392956959195081e-05,
1075
+ "loss": 0.5941,
1076
+ "step": 1250
1077
+ },
1078
+ {
1079
+ "epoch": 0.33985861881457313,
1080
+ "eval_loss": 0.5951874852180481,
1081
+ "eval_runtime": 101.3262,
1082
+ "eval_samples_per_second": 4.935,
1083
+ "eval_steps_per_second": 0.158,
1084
+ "step": 1250
1085
+ },
1086
+ {
1087
+ "epoch": 0.3425774877650897,
1088
+ "grad_norm": 0.314453125,
1089
+ "learning_rate": 3.378982671883734e-05,
1090
+ "loss": 0.6055,
1091
+ "step": 1260
1092
+ },
1093
+ {
1094
+ "epoch": 0.3452963567156063,
1095
+ "grad_norm": 0.296875,
1096
+ "learning_rate": 3.3650083845723866e-05,
1097
+ "loss": 0.6074,
1098
+ "step": 1270
1099
+ },
1100
+ {
1101
+ "epoch": 0.3480152256661229,
1102
+ "grad_norm": 0.2578125,
1103
+ "learning_rate": 3.35103409726104e-05,
1104
+ "loss": 0.593,
1105
+ "step": 1280
1106
+ },
1107
+ {
1108
+ "epoch": 0.35073409461663946,
1109
+ "grad_norm": 0.291015625,
1110
+ "learning_rate": 3.3370598099496925e-05,
1111
+ "loss": 0.6051,
1112
+ "step": 1290
1113
+ },
1114
+ {
1115
+ "epoch": 0.35345296356715605,
1116
+ "grad_norm": 0.29296875,
1117
+ "learning_rate": 3.323085522638346e-05,
1118
+ "loss": 0.5938,
1119
+ "step": 1300
1120
+ },
1121
+ {
1122
+ "epoch": 0.35345296356715605,
1123
+ "eval_loss": 0.594124972820282,
1124
+ "eval_runtime": 99.7797,
1125
+ "eval_samples_per_second": 5.011,
1126
+ "eval_steps_per_second": 0.16,
1127
+ "step": 1300
1128
+ },
1129
+ {
1130
+ "epoch": 0.35617183251767265,
1131
+ "grad_norm": 0.26953125,
1132
+ "learning_rate": 3.309111235326999e-05,
1133
+ "loss": 0.6051,
1134
+ "step": 1310
1135
+ },
1136
+ {
1137
+ "epoch": 0.35889070146818924,
1138
+ "grad_norm": 0.29296875,
1139
+ "learning_rate": 3.2951369480156516e-05,
1140
+ "loss": 0.5957,
1141
+ "step": 1320
1142
+ },
1143
+ {
1144
+ "epoch": 0.36160957041870584,
1145
+ "grad_norm": 0.25390625,
1146
+ "learning_rate": 3.281162660704304e-05,
1147
+ "loss": 0.5961,
1148
+ "step": 1330
1149
+ },
1150
+ {
1151
+ "epoch": 0.3643284393692224,
1152
+ "grad_norm": 0.29296875,
1153
+ "learning_rate": 3.267188373392957e-05,
1154
+ "loss": 0.6035,
1155
+ "step": 1340
1156
+ },
1157
+ {
1158
+ "epoch": 0.367047308319739,
1159
+ "grad_norm": 0.271484375,
1160
+ "learning_rate": 3.25321408608161e-05,
1161
+ "loss": 0.5977,
1162
+ "step": 1350
1163
+ },
1164
+ {
1165
+ "epoch": 0.367047308319739,
1166
+ "eval_loss": 0.5958437323570251,
1167
+ "eval_runtime": 99.1382,
1168
+ "eval_samples_per_second": 5.043,
1169
+ "eval_steps_per_second": 0.161,
1170
+ "step": 1350
1171
+ },
1172
+ {
1173
+ "epoch": 0.36976617727025557,
1174
+ "grad_norm": 0.255859375,
1175
+ "learning_rate": 3.2392397987702626e-05,
1176
+ "loss": 0.6008,
1177
+ "step": 1360
1178
+ },
1179
+ {
1180
+ "epoch": 0.37248504622077216,
1181
+ "grad_norm": 0.29296875,
1182
+ "learning_rate": 3.225265511458916e-05,
1183
+ "loss": 0.5977,
1184
+ "step": 1370
1185
+ },
1186
+ {
1187
+ "epoch": 0.37520391517128876,
1188
+ "grad_norm": 0.291015625,
1189
+ "learning_rate": 3.2112912241475685e-05,
1190
+ "loss": 0.6012,
1191
+ "step": 1380
1192
+ },
1193
+ {
1194
+ "epoch": 0.37792278412180536,
1195
+ "grad_norm": 0.25,
1196
+ "learning_rate": 3.197316936836222e-05,
1197
+ "loss": 0.5949,
1198
+ "step": 1390
1199
+ },
1200
+ {
1201
+ "epoch": 0.3806416530723219,
1202
+ "grad_norm": 0.2197265625,
1203
+ "learning_rate": 3.183342649524874e-05,
1204
+ "loss": 0.6004,
1205
+ "step": 1400
1206
+ },
1207
+ {
1208
+ "epoch": 0.3806416530723219,
1209
+ "eval_loss": 0.5949375033378601,
1210
+ "eval_runtime": 123.9662,
1211
+ "eval_samples_per_second": 4.033,
1212
+ "eval_steps_per_second": 0.129,
1213
+ "step": 1400
1214
+ },
1215
+ {
1216
+ "epoch": 0.3833605220228385,
1217
+ "grad_norm": 0.2734375,
1218
+ "learning_rate": 3.1693683622135276e-05,
1219
+ "loss": 0.598,
1220
+ "step": 1410
1221
+ },
1222
+ {
1223
+ "epoch": 0.3860793909733551,
1224
+ "grad_norm": 0.490234375,
1225
+ "learning_rate": 3.15539407490218e-05,
1226
+ "loss": 0.5934,
1227
+ "step": 1420
1228
+ },
1229
+ {
1230
+ "epoch": 0.3887982599238717,
1231
+ "grad_norm": 0.27734375,
1232
+ "learning_rate": 3.141419787590833e-05,
1233
+ "loss": 0.6004,
1234
+ "step": 1430
1235
+ },
1236
+ {
1237
+ "epoch": 0.3915171288743883,
1238
+ "grad_norm": 0.25390625,
1239
+ "learning_rate": 3.127445500279486e-05,
1240
+ "loss": 0.5969,
1241
+ "step": 1440
1242
+ },
1243
+ {
1244
+ "epoch": 0.3942359978249048,
1245
+ "grad_norm": 0.25,
1246
+ "learning_rate": 3.1134712129681386e-05,
1247
+ "loss": 0.6012,
1248
+ "step": 1450
1249
+ },
1250
+ {
1251
+ "epoch": 0.3942359978249048,
1252
+ "eval_loss": 0.5909374952316284,
1253
+ "eval_runtime": 110.1458,
1254
+ "eval_samples_per_second": 4.539,
1255
+ "eval_steps_per_second": 0.145,
1256
+ "step": 1450
1257
+ },
1258
+ {
1259
+ "epoch": 0.3969548667754214,
1260
+ "grad_norm": 0.2470703125,
1261
+ "learning_rate": 3.099496925656792e-05,
1262
+ "loss": 0.607,
1263
+ "step": 1460
1264
+ },
1265
+ {
1266
+ "epoch": 0.399673735725938,
1267
+ "grad_norm": 0.37109375,
1268
+ "learning_rate": 3.0855226383454444e-05,
1269
+ "loss": 0.6055,
1270
+ "step": 1470
1271
+ },
1272
+ {
1273
+ "epoch": 0.4023926046764546,
1274
+ "grad_norm": 0.259765625,
1275
+ "learning_rate": 3.071548351034098e-05,
1276
+ "loss": 0.5961,
1277
+ "step": 1480
1278
+ },
1279
+ {
1280
+ "epoch": 0.4051114736269712,
1281
+ "grad_norm": 0.287109375,
1282
+ "learning_rate": 3.05757406372275e-05,
1283
+ "loss": 0.5957,
1284
+ "step": 1490
1285
+ },
1286
+ {
1287
+ "epoch": 0.4078303425774878,
1288
+ "grad_norm": 0.30078125,
1289
+ "learning_rate": 3.0435997764114032e-05,
1290
+ "loss": 0.5957,
1291
+ "step": 1500
1292
+ },
1293
+ {
1294
+ "epoch": 0.4078303425774878,
1295
+ "eval_loss": 0.5928750038146973,
1296
+ "eval_runtime": 105.1416,
1297
+ "eval_samples_per_second": 4.755,
1298
+ "eval_steps_per_second": 0.152,
1299
+ "step": 1500
1300
+ },
1301
+ {
1302
+ "epoch": 0.41054921152800433,
1303
+ "grad_norm": 0.25390625,
1304
+ "learning_rate": 3.0296254891000558e-05,
1305
+ "loss": 0.6062,
1306
+ "step": 1510
1307
+ },
1308
+ {
1309
+ "epoch": 0.4132680804785209,
1310
+ "grad_norm": 0.259765625,
1311
+ "learning_rate": 3.015651201788709e-05,
1312
+ "loss": 0.5965,
1313
+ "step": 1520
1314
+ },
1315
+ {
1316
+ "epoch": 0.4159869494290375,
1317
+ "grad_norm": 0.29296875,
1318
+ "learning_rate": 3.0016769144773617e-05,
1319
+ "loss": 0.6125,
1320
+ "step": 1530
1321
+ },
1322
+ {
1323
+ "epoch": 0.4187058183795541,
1324
+ "grad_norm": 0.265625,
1325
+ "learning_rate": 2.9877026271660146e-05,
1326
+ "loss": 0.5992,
1327
+ "step": 1540
1328
+ },
1329
+ {
1330
+ "epoch": 0.4214246873300707,
1331
+ "grad_norm": 0.27734375,
1332
+ "learning_rate": 2.973728339854668e-05,
1333
+ "loss": 0.5961,
1334
+ "step": 1550
1335
+ },
1336
+ {
1337
+ "epoch": 0.4214246873300707,
1338
+ "eval_loss": 0.5925312638282776,
1339
+ "eval_runtime": 104.3951,
1340
+ "eval_samples_per_second": 4.789,
1341
+ "eval_steps_per_second": 0.153,
1342
+ "step": 1550
1343
+ },
1344
+ {
1345
+ "epoch": 0.42414355628058725,
1346
+ "grad_norm": 0.31640625,
1347
+ "learning_rate": 2.9597540525433204e-05,
1348
+ "loss": 0.598,
1349
+ "step": 1560
1350
+ },
1351
+ {
1352
+ "epoch": 0.42686242523110385,
1353
+ "grad_norm": 0.265625,
1354
+ "learning_rate": 2.9457797652319734e-05,
1355
+ "loss": 0.5953,
1356
+ "step": 1570
1357
+ },
1358
+ {
1359
+ "epoch": 0.42958129418162044,
1360
+ "grad_norm": 0.26953125,
1361
+ "learning_rate": 2.931805477920626e-05,
1362
+ "loss": 0.6012,
1363
+ "step": 1580
1364
+ },
1365
+ {
1366
+ "epoch": 0.43230016313213704,
1367
+ "grad_norm": 0.271484375,
1368
+ "learning_rate": 2.9178311906092792e-05,
1369
+ "loss": 0.6035,
1370
+ "step": 1590
1371
+ },
1372
+ {
1373
+ "epoch": 0.43501903208265363,
1374
+ "grad_norm": 0.380859375,
1375
+ "learning_rate": 2.9038569032979318e-05,
1376
+ "loss": 0.593,
1377
+ "step": 1600
1378
+ },
1379
+ {
1380
+ "epoch": 0.43501903208265363,
1381
+ "eval_loss": 0.5901250243186951,
1382
+ "eval_runtime": 104.8022,
1383
+ "eval_samples_per_second": 4.771,
1384
+ "eval_steps_per_second": 0.153,
1385
+ "step": 1600
1386
+ },
1387
+ {
1388
+ "epoch": 0.4377379010331702,
1389
+ "grad_norm": 0.283203125,
1390
+ "learning_rate": 2.889882615986585e-05,
1391
+ "loss": 0.602,
1392
+ "step": 1610
1393
+ },
1394
+ {
1395
+ "epoch": 0.44045676998368677,
1396
+ "grad_norm": 0.41015625,
1397
+ "learning_rate": 2.8759083286752376e-05,
1398
+ "loss": 0.6004,
1399
+ "step": 1620
1400
+ },
1401
+ {
1402
+ "epoch": 0.44317563893420336,
1403
+ "grad_norm": 0.30078125,
1404
+ "learning_rate": 2.8619340413638906e-05,
1405
+ "loss": 0.5984,
1406
+ "step": 1630
1407
+ },
1408
+ {
1409
+ "epoch": 0.44589450788471996,
1410
+ "grad_norm": 0.28125,
1411
+ "learning_rate": 2.847959754052543e-05,
1412
+ "loss": 0.5863,
1413
+ "step": 1640
1414
+ },
1415
+ {
1416
+ "epoch": 0.44861337683523655,
1417
+ "grad_norm": 0.25,
1418
+ "learning_rate": 2.8339854667411964e-05,
1419
+ "loss": 0.5906,
1420
+ "step": 1650
1421
+ },
1422
+ {
1423
+ "epoch": 0.44861337683523655,
1424
+ "eval_loss": 0.5921249985694885,
1425
+ "eval_runtime": 102.6549,
1426
+ "eval_samples_per_second": 4.871,
1427
+ "eval_steps_per_second": 0.156,
1428
+ "step": 1650
1429
+ },
1430
+ {
1431
+ "epoch": 0.45133224578575315,
1432
+ "grad_norm": 0.294921875,
1433
+ "learning_rate": 2.820011179429849e-05,
1434
+ "loss": 0.6012,
1435
+ "step": 1660
1436
+ },
1437
+ {
1438
+ "epoch": 0.4540511147362697,
1439
+ "grad_norm": 0.27734375,
1440
+ "learning_rate": 2.806036892118502e-05,
1441
+ "loss": 0.5906,
1442
+ "step": 1670
1443
+ },
1444
+ {
1445
+ "epoch": 0.4567699836867863,
1446
+ "grad_norm": 0.310546875,
1447
+ "learning_rate": 2.7920626048071552e-05,
1448
+ "loss": 0.5992,
1449
+ "step": 1680
1450
+ },
1451
+ {
1452
+ "epoch": 0.4594888526373029,
1453
+ "grad_norm": 0.275390625,
1454
+ "learning_rate": 2.7780883174958078e-05,
1455
+ "loss": 0.6012,
1456
+ "step": 1690
1457
+ },
1458
+ {
1459
+ "epoch": 0.4622077215878195,
1460
+ "grad_norm": 0.296875,
1461
+ "learning_rate": 2.764114030184461e-05,
1462
+ "loss": 0.5926,
1463
+ "step": 1700
1464
+ },
1465
+ {
1466
+ "epoch": 0.4622077215878195,
1467
+ "eval_loss": 0.5884062647819519,
1468
+ "eval_runtime": 103.7651,
1469
+ "eval_samples_per_second": 4.819,
1470
+ "eval_steps_per_second": 0.154,
1471
+ "step": 1700
1472
+ },
1473
+ {
1474
+ "epoch": 0.46492659053833607,
1475
+ "grad_norm": 0.275390625,
1476
+ "learning_rate": 2.7501397428731136e-05,
1477
+ "loss": 0.6012,
1478
+ "step": 1710
1479
+ },
1480
+ {
1481
+ "epoch": 0.4676454594888526,
1482
+ "grad_norm": 0.2890625,
1483
+ "learning_rate": 2.7361654555617665e-05,
1484
+ "loss": 0.6039,
1485
+ "step": 1720
1486
+ },
1487
+ {
1488
+ "epoch": 0.4703643284393692,
1489
+ "grad_norm": 0.2890625,
1490
+ "learning_rate": 2.722191168250419e-05,
1491
+ "loss": 0.5938,
1492
+ "step": 1730
1493
+ },
1494
+ {
1495
+ "epoch": 0.4730831973898858,
1496
+ "grad_norm": 0.416015625,
1497
+ "learning_rate": 2.7082168809390724e-05,
1498
+ "loss": 0.5984,
1499
+ "step": 1740
1500
+ },
1501
+ {
1502
+ "epoch": 0.4758020663404024,
1503
+ "grad_norm": 0.486328125,
1504
+ "learning_rate": 2.694242593627725e-05,
1505
+ "loss": 0.5926,
1506
+ "step": 1750
1507
+ },
1508
+ {
1509
+ "epoch": 0.4758020663404024,
1510
+ "eval_loss": 0.5889062285423279,
1511
+ "eval_runtime": 100.2362,
1512
+ "eval_samples_per_second": 4.988,
1513
+ "eval_steps_per_second": 0.16,
1514
+ "step": 1750
1515
+ },
1516
+ {
1517
+ "epoch": 0.478520935290919,
1518
+ "grad_norm": 0.291015625,
1519
+ "learning_rate": 2.680268306316378e-05,
1520
+ "loss": 0.5984,
1521
+ "step": 1760
1522
+ },
1523
+ {
1524
+ "epoch": 0.4812398042414356,
1525
+ "grad_norm": 0.2578125,
1526
+ "learning_rate": 2.6662940190050305e-05,
1527
+ "loss": 0.5977,
1528
+ "step": 1770
1529
+ },
1530
+ {
1531
+ "epoch": 0.4839586731919521,
1532
+ "grad_norm": 0.330078125,
1533
+ "learning_rate": 2.6523197316936838e-05,
1534
+ "loss": 0.5957,
1535
+ "step": 1780
1536
+ },
1537
+ {
1538
+ "epoch": 0.4866775421424687,
1539
+ "grad_norm": 0.3046875,
1540
+ "learning_rate": 2.638345444382337e-05,
1541
+ "loss": 0.5918,
1542
+ "step": 1790
1543
+ },
1544
+ {
1545
+ "epoch": 0.4893964110929853,
1546
+ "grad_norm": 0.30859375,
1547
+ "learning_rate": 2.6243711570709896e-05,
1548
+ "loss": 0.6027,
1549
+ "step": 1800
1550
+ },
1551
+ {
1552
+ "epoch": 0.4893964110929853,
1553
+ "eval_loss": 0.5900312662124634,
1554
+ "eval_runtime": 101.8319,
1555
+ "eval_samples_per_second": 4.91,
1556
+ "eval_steps_per_second": 0.157,
1557
+ "step": 1800
1558
+ },
1559
+ {
1560
+ "epoch": 0.4921152800435019,
1561
+ "grad_norm": 0.287109375,
1562
+ "learning_rate": 2.6103968697596425e-05,
1563
+ "loss": 0.5973,
1564
+ "step": 1810
1565
+ },
1566
+ {
1567
+ "epoch": 0.4948341489940185,
1568
+ "grad_norm": 0.275390625,
1569
+ "learning_rate": 2.596422582448295e-05,
1570
+ "loss": 0.591,
1571
+ "step": 1820
1572
+ },
1573
+ {
1574
+ "epoch": 0.49755301794453505,
1575
+ "grad_norm": 0.2890625,
1576
+ "learning_rate": 2.5824482951369484e-05,
1577
+ "loss": 0.5941,
1578
+ "step": 1830
1579
+ },
1580
+ {
1581
+ "epoch": 0.5002718868950516,
1582
+ "grad_norm": 0.326171875,
1583
+ "learning_rate": 2.568474007825601e-05,
1584
+ "loss": 0.5945,
1585
+ "step": 1840
1586
+ },
1587
+ {
1588
+ "epoch": 0.5029907558455683,
1589
+ "grad_norm": 0.328125,
1590
+ "learning_rate": 2.554499720514254e-05,
1591
+ "loss": 0.5961,
1592
+ "step": 1850
1593
+ },
1594
+ {
1595
+ "epoch": 0.5029907558455683,
1596
+ "eval_loss": 0.5892187356948853,
1597
+ "eval_runtime": 100.0652,
1598
+ "eval_samples_per_second": 4.997,
1599
+ "eval_steps_per_second": 0.16,
1600
+ "step": 1850
1601
+ },
1602
+ {
1603
+ "epoch": 0.5057096247960848,
1604
+ "grad_norm": 0.26953125,
1605
+ "learning_rate": 2.5405254332029065e-05,
1606
+ "loss": 0.5895,
1607
+ "step": 1860
1608
+ },
1609
+ {
1610
+ "epoch": 0.5084284937466014,
1611
+ "grad_norm": 0.28515625,
1612
+ "learning_rate": 2.5265511458915597e-05,
1613
+ "loss": 0.5906,
1614
+ "step": 1870
1615
+ },
1616
+ {
1617
+ "epoch": 0.511147362697118,
1618
+ "grad_norm": 0.31640625,
1619
+ "learning_rate": 2.5125768585802123e-05,
1620
+ "loss": 0.5906,
1621
+ "step": 1880
1622
+ },
1623
+ {
1624
+ "epoch": 0.5138662316476346,
1625
+ "grad_norm": 0.2890625,
1626
+ "learning_rate": 2.4986025712688656e-05,
1627
+ "loss": 0.6008,
1628
+ "step": 1890
1629
+ },
1630
+ {
1631
+ "epoch": 0.5165851005981512,
1632
+ "grad_norm": 0.294921875,
1633
+ "learning_rate": 2.484628283957518e-05,
1634
+ "loss": 0.5906,
1635
+ "step": 1900
1636
+ },
1637
+ {
1638
+ "epoch": 0.5165851005981512,
1639
+ "eval_loss": 0.5868750214576721,
1640
+ "eval_runtime": 110.2559,
1641
+ "eval_samples_per_second": 4.535,
1642
+ "eval_steps_per_second": 0.145,
1643
+ "step": 1900
1644
+ },
1645
+ {
1646
+ "epoch": 0.5193039695486678,
1647
+ "grad_norm": 0.30078125,
1648
+ "learning_rate": 2.470653996646171e-05,
1649
+ "loss": 0.5938,
1650
+ "step": 1910
1651
+ },
1652
+ {
1653
+ "epoch": 0.5220228384991843,
1654
+ "grad_norm": 0.275390625,
1655
+ "learning_rate": 2.456679709334824e-05,
1656
+ "loss": 0.5887,
1657
+ "step": 1920
1658
+ },
1659
+ {
1660
+ "epoch": 0.524741707449701,
1661
+ "grad_norm": 0.3125,
1662
+ "learning_rate": 2.442705422023477e-05,
1663
+ "loss": 0.593,
1664
+ "step": 1930
1665
+ },
1666
+ {
1667
+ "epoch": 0.5274605764002175,
1668
+ "grad_norm": 0.287109375,
1669
+ "learning_rate": 2.42873113471213e-05,
1670
+ "loss": 0.5996,
1671
+ "step": 1940
1672
+ },
1673
+ {
1674
+ "epoch": 0.5301794453507341,
1675
+ "grad_norm": 0.279296875,
1676
+ "learning_rate": 2.4147568474007825e-05,
1677
+ "loss": 0.6016,
1678
+ "step": 1950
1679
+ },
1680
+ {
1681
+ "epoch": 0.5301794453507341,
1682
+ "eval_loss": 0.5910937786102295,
1683
+ "eval_runtime": 105.3606,
1684
+ "eval_samples_per_second": 4.746,
1685
+ "eval_steps_per_second": 0.152,
1686
+ "step": 1950
1687
+ },
1688
+ {
1689
+ "epoch": 0.5328983143012507,
1690
+ "grad_norm": 0.2890625,
1691
+ "learning_rate": 2.4007825600894354e-05,
1692
+ "loss": 0.5977,
1693
+ "step": 1960
1694
+ },
1695
+ {
1696
+ "epoch": 0.5356171832517672,
1697
+ "grad_norm": 0.296875,
1698
+ "learning_rate": 2.3868082727780886e-05,
1699
+ "loss": 0.5977,
1700
+ "step": 1970
1701
+ },
1702
+ {
1703
+ "epoch": 0.5383360522022839,
1704
+ "grad_norm": 0.40234375,
1705
+ "learning_rate": 2.3728339854667416e-05,
1706
+ "loss": 0.5961,
1707
+ "step": 1980
1708
+ },
1709
+ {
1710
+ "epoch": 0.5410549211528004,
1711
+ "grad_norm": 0.302734375,
1712
+ "learning_rate": 2.358859698155394e-05,
1713
+ "loss": 0.5969,
1714
+ "step": 1990
1715
+ },
1716
+ {
1717
+ "epoch": 0.5437737901033171,
1718
+ "grad_norm": 0.412109375,
1719
+ "learning_rate": 2.344885410844047e-05,
1720
+ "loss": 0.5895,
1721
+ "step": 2000
1722
+ },
1723
+ {
1724
+ "epoch": 0.5437737901033171,
1725
+ "eval_loss": 0.5891249775886536,
1726
+ "eval_runtime": 103.6626,
1727
+ "eval_samples_per_second": 4.823,
1728
+ "eval_steps_per_second": 0.154,
1729
+ "step": 2000
1730
+ },
1731
+ {
1732
+ "epoch": 0.5464926590538336,
1733
+ "grad_norm": 0.294921875,
1734
+ "learning_rate": 2.3309111235327e-05,
1735
+ "loss": 0.6016,
1736
+ "step": 2010
1737
+ },
1738
+ {
1739
+ "epoch": 0.5492115280043501,
1740
+ "grad_norm": 0.29296875,
1741
+ "learning_rate": 2.316936836221353e-05,
1742
+ "loss": 0.6074,
1743
+ "step": 2020
1744
+ },
1745
+ {
1746
+ "epoch": 0.5519303969548668,
1747
+ "grad_norm": 0.283203125,
1748
+ "learning_rate": 2.302962548910006e-05,
1749
+ "loss": 0.5875,
1750
+ "step": 2030
1751
+ },
1752
+ {
1753
+ "epoch": 0.5546492659053833,
1754
+ "grad_norm": 0.484375,
1755
+ "learning_rate": 2.2889882615986584e-05,
1756
+ "loss": 0.5984,
1757
+ "step": 2040
1758
+ },
1759
+ {
1760
+ "epoch": 0.5573681348559,
1761
+ "grad_norm": 0.30859375,
1762
+ "learning_rate": 2.2750139742873114e-05,
1763
+ "loss": 0.5906,
1764
+ "step": 2050
1765
+ },
1766
+ {
1767
+ "epoch": 0.5573681348559,
1768
+ "eval_loss": 0.5882187485694885,
1769
+ "eval_runtime": 103.7114,
1770
+ "eval_samples_per_second": 4.821,
1771
+ "eval_steps_per_second": 0.154,
1772
+ "step": 2050
1773
+ },
1774
+ {
1775
+ "epoch": 0.5600870038064165,
1776
+ "grad_norm": 0.291015625,
1777
+ "learning_rate": 2.2610396869759643e-05,
1778
+ "loss": 0.5969,
1779
+ "step": 2060
1780
+ },
1781
+ {
1782
+ "epoch": 0.5628058727569332,
1783
+ "grad_norm": 0.291015625,
1784
+ "learning_rate": 2.2470653996646172e-05,
1785
+ "loss": 0.5859,
1786
+ "step": 2070
1787
+ },
1788
+ {
1789
+ "epoch": 0.5655247417074497,
1790
+ "grad_norm": 0.318359375,
1791
+ "learning_rate": 2.2330911123532698e-05,
1792
+ "loss": 0.6016,
1793
+ "step": 2080
1794
+ },
1795
+ {
1796
+ "epoch": 0.5682436106579662,
1797
+ "grad_norm": 0.26953125,
1798
+ "learning_rate": 2.219116825041923e-05,
1799
+ "loss": 0.5977,
1800
+ "step": 2090
1801
+ },
1802
+ {
1803
+ "epoch": 0.5709624796084829,
1804
+ "grad_norm": 0.271484375,
1805
+ "learning_rate": 2.205142537730576e-05,
1806
+ "loss": 0.5926,
1807
+ "step": 2100
1808
+ },
1809
+ {
1810
+ "epoch": 0.5709624796084829,
1811
+ "eval_loss": 0.5868750214576721,
1812
+ "eval_runtime": 102.6198,
1813
+ "eval_samples_per_second": 4.872,
1814
+ "eval_steps_per_second": 0.156,
1815
+ "step": 2100
1816
+ },
1817
+ {
1818
+ "epoch": 0.5736813485589994,
1819
+ "grad_norm": 0.39453125,
1820
+ "learning_rate": 2.191168250419229e-05,
1821
+ "loss": 0.5957,
1822
+ "step": 2110
1823
+ },
1824
+ {
1825
+ "epoch": 0.5764002175095161,
1826
+ "grad_norm": 0.29296875,
1827
+ "learning_rate": 2.1771939631078815e-05,
1828
+ "loss": 0.591,
1829
+ "step": 2120
1830
+ },
1831
+ {
1832
+ "epoch": 0.5791190864600326,
1833
+ "grad_norm": 0.279296875,
1834
+ "learning_rate": 2.1632196757965344e-05,
1835
+ "loss": 0.5918,
1836
+ "step": 2130
1837
+ },
1838
+ {
1839
+ "epoch": 0.5818379554105492,
1840
+ "grad_norm": 0.2890625,
1841
+ "learning_rate": 2.1492453884851873e-05,
1842
+ "loss": 0.5949,
1843
+ "step": 2140
1844
+ },
1845
+ {
1846
+ "epoch": 0.5845568243610658,
1847
+ "grad_norm": 0.341796875,
1848
+ "learning_rate": 2.1352711011738403e-05,
1849
+ "loss": 0.5922,
1850
+ "step": 2150
1851
+ },
1852
+ {
1853
+ "epoch": 0.5845568243610658,
1854
+ "eval_loss": 0.5886250138282776,
1855
+ "eval_runtime": 101.2748,
1856
+ "eval_samples_per_second": 4.937,
1857
+ "eval_steps_per_second": 0.158,
1858
+ "step": 2150
1859
+ },
1860
+ {
1861
+ "epoch": 0.5872756933115824,
1862
+ "grad_norm": 0.291015625,
1863
+ "learning_rate": 2.1212968138624932e-05,
1864
+ "loss": 0.5887,
1865
+ "step": 2160
1866
+ },
1867
+ {
1868
+ "epoch": 0.589994562262099,
1869
+ "grad_norm": 0.298828125,
1870
+ "learning_rate": 2.1073225265511458e-05,
1871
+ "loss": 0.5949,
1872
+ "step": 2170
1873
+ },
1874
+ {
1875
+ "epoch": 0.5927134312126155,
1876
+ "grad_norm": 0.3046875,
1877
+ "learning_rate": 2.0933482392397987e-05,
1878
+ "loss": 0.593,
1879
+ "step": 2180
1880
+ },
1881
+ {
1882
+ "epoch": 0.5954323001631321,
1883
+ "grad_norm": 0.265625,
1884
+ "learning_rate": 2.0793739519284516e-05,
1885
+ "loss": 0.5977,
1886
+ "step": 2190
1887
+ },
1888
+ {
1889
+ "epoch": 0.5981511691136487,
1890
+ "grad_norm": 0.291015625,
1891
+ "learning_rate": 2.0653996646171045e-05,
1892
+ "loss": 0.591,
1893
+ "step": 2200
1894
+ },
1895
+ {
1896
+ "epoch": 0.5981511691136487,
1897
+ "eval_loss": 0.5874687433242798,
1898
+ "eval_runtime": 99.4212,
1899
+ "eval_samples_per_second": 5.029,
1900
+ "eval_steps_per_second": 0.161,
1901
+ "step": 2200
1902
+ },
1903
+ {
1904
+ "epoch": 0.6008700380641653,
1905
+ "grad_norm": 0.291015625,
1906
+ "learning_rate": 2.0514253773057575e-05,
1907
+ "loss": 0.5945,
1908
+ "step": 2210
1909
+ },
1910
+ {
1911
+ "epoch": 0.6035889070146819,
1912
+ "grad_norm": 0.283203125,
1913
+ "learning_rate": 2.0374510899944104e-05,
1914
+ "loss": 0.5961,
1915
+ "step": 2220
1916
+ },
1917
+ {
1918
+ "epoch": 0.6063077759651985,
1919
+ "grad_norm": 0.283203125,
1920
+ "learning_rate": 2.0234768026830633e-05,
1921
+ "loss": 0.5918,
1922
+ "step": 2230
1923
+ },
1924
+ {
1925
+ "epoch": 0.609026644915715,
1926
+ "grad_norm": 0.32421875,
1927
+ "learning_rate": 2.0095025153717162e-05,
1928
+ "loss": 0.5918,
1929
+ "step": 2240
1930
+ },
1931
+ {
1932
+ "epoch": 0.6117455138662317,
1933
+ "grad_norm": 0.27734375,
1934
+ "learning_rate": 1.995528228060369e-05,
1935
+ "loss": 0.5879,
1936
+ "step": 2250
1937
+ },
1938
+ {
1939
+ "epoch": 0.6117455138662317,
1940
+ "eval_loss": 0.5858437418937683,
1941
+ "eval_runtime": 99.61,
1942
+ "eval_samples_per_second": 5.02,
1943
+ "eval_steps_per_second": 0.161,
1944
+ "step": 2250
1945
+ },
1946
+ {
1947
+ "epoch": 0.6144643828167482,
1948
+ "grad_norm": 0.25390625,
1949
+ "learning_rate": 1.9815539407490218e-05,
1950
+ "loss": 0.5957,
1951
+ "step": 2260
1952
+ },
1953
+ {
1954
+ "epoch": 0.6171832517672649,
1955
+ "grad_norm": 0.2734375,
1956
+ "learning_rate": 1.9675796534376747e-05,
1957
+ "loss": 0.584,
1958
+ "step": 2270
1959
+ },
1960
+ {
1961
+ "epoch": 0.6199021207177814,
1962
+ "grad_norm": 0.310546875,
1963
+ "learning_rate": 1.9536053661263276e-05,
1964
+ "loss": 0.5934,
1965
+ "step": 2280
1966
+ },
1967
+ {
1968
+ "epoch": 0.622620989668298,
1969
+ "grad_norm": 0.33984375,
1970
+ "learning_rate": 1.9396310788149805e-05,
1971
+ "loss": 0.5902,
1972
+ "step": 2290
1973
+ },
1974
+ {
1975
+ "epoch": 0.6253398586188146,
1976
+ "grad_norm": 0.328125,
1977
+ "learning_rate": 1.9256567915036335e-05,
1978
+ "loss": 0.591,
1979
+ "step": 2300
1980
+ },
1981
+ {
1982
+ "epoch": 0.6253398586188146,
1983
+ "eval_loss": 0.5860312581062317,
1984
+ "eval_runtime": 99.6003,
1985
+ "eval_samples_per_second": 5.02,
1986
+ "eval_steps_per_second": 0.161,
1987
+ "step": 2300
1988
+ },
1989
+ {
1990
+ "epoch": 0.6280587275693311,
1991
+ "grad_norm": 0.3046875,
1992
+ "learning_rate": 1.911682504192286e-05,
1993
+ "loss": 0.5938,
1994
+ "step": 2310
1995
+ },
1996
+ {
1997
+ "epoch": 0.6307775965198478,
1998
+ "grad_norm": 0.294921875,
1999
+ "learning_rate": 1.897708216880939e-05,
2000
+ "loss": 0.5945,
2001
+ "step": 2320
2002
+ },
2003
+ {
2004
+ "epoch": 0.6334964654703643,
2005
+ "grad_norm": 0.30859375,
2006
+ "learning_rate": 1.8837339295695922e-05,
2007
+ "loss": 0.5906,
2008
+ "step": 2330
2009
+ },
2010
+ {
2011
+ "epoch": 0.636215334420881,
2012
+ "grad_norm": 0.28125,
2013
+ "learning_rate": 1.869759642258245e-05,
2014
+ "loss": 0.5961,
2015
+ "step": 2340
2016
+ },
2017
+ {
2018
+ "epoch": 0.6389342033713975,
2019
+ "grad_norm": 0.306640625,
2020
+ "learning_rate": 1.8557853549468977e-05,
2021
+ "loss": 0.5906,
2022
+ "step": 2350
2023
+ },
2024
+ {
2025
+ "epoch": 0.6389342033713975,
2026
+ "eval_loss": 0.5859687328338623,
2027
+ "eval_runtime": 98.5084,
2028
+ "eval_samples_per_second": 5.076,
2029
+ "eval_steps_per_second": 0.162,
2030
+ "step": 2350
2031
+ },
2032
+ {
2033
+ "epoch": 0.641653072321914,
2034
+ "grad_norm": 0.283203125,
2035
+ "learning_rate": 1.8418110676355507e-05,
2036
+ "loss": 0.5887,
2037
+ "step": 2360
2038
+ },
2039
+ {
2040
+ "epoch": 0.6443719412724307,
2041
+ "grad_norm": 0.27734375,
2042
+ "learning_rate": 1.8278367803242036e-05,
2043
+ "loss": 0.5887,
2044
+ "step": 2370
2045
+ },
2046
+ {
2047
+ "epoch": 0.6470908102229472,
2048
+ "grad_norm": 0.306640625,
2049
+ "learning_rate": 1.8138624930128565e-05,
2050
+ "loss": 0.5867,
2051
+ "step": 2380
2052
+ },
2053
+ {
2054
+ "epoch": 0.6498096791734639,
2055
+ "grad_norm": 0.35546875,
2056
+ "learning_rate": 1.7998882057015094e-05,
2057
+ "loss": 0.593,
2058
+ "step": 2390
2059
+ },
2060
+ {
2061
+ "epoch": 0.6525285481239804,
2062
+ "grad_norm": 0.306640625,
2063
+ "learning_rate": 1.785913918390162e-05,
2064
+ "loss": 0.5977,
2065
+ "step": 2400
2066
+ },
2067
+ {
2068
+ "epoch": 0.6525285481239804,
2069
+ "eval_loss": 0.5853124856948853,
2070
+ "eval_runtime": 116.5368,
2071
+ "eval_samples_per_second": 4.29,
2072
+ "eval_steps_per_second": 0.137,
2073
+ "step": 2400
2074
+ },
2075
+ {
2076
+ "epoch": 0.655247417074497,
2077
+ "grad_norm": 0.29296875,
2078
+ "learning_rate": 1.771939631078815e-05,
2079
+ "loss": 0.5953,
2080
+ "step": 2410
2081
+ },
2082
+ {
2083
+ "epoch": 0.6579662860250136,
2084
+ "grad_norm": 0.3984375,
2085
+ "learning_rate": 1.757965343767468e-05,
2086
+ "loss": 0.5875,
2087
+ "step": 2420
2088
+ },
2089
+ {
2090
+ "epoch": 0.6606851549755302,
2091
+ "grad_norm": 0.330078125,
2092
+ "learning_rate": 1.7439910564561208e-05,
2093
+ "loss": 0.5824,
2094
+ "step": 2430
2095
+ },
2096
+ {
2097
+ "epoch": 0.6634040239260468,
2098
+ "grad_norm": 0.57421875,
2099
+ "learning_rate": 1.7300167691447737e-05,
2100
+ "loss": 0.5902,
2101
+ "step": 2440
2102
+ },
2103
+ {
2104
+ "epoch": 0.6661228928765633,
2105
+ "grad_norm": 0.38671875,
2106
+ "learning_rate": 1.7160424818334266e-05,
2107
+ "loss": 0.5926,
2108
+ "step": 2450
2109
+ },
2110
+ {
2111
+ "epoch": 0.6661228928765633,
2112
+ "eval_loss": 0.5852187275886536,
2113
+ "eval_runtime": 103.4693,
2114
+ "eval_samples_per_second": 4.832,
2115
+ "eval_steps_per_second": 0.155,
2116
+ "step": 2450
2117
+ },
2118
+ {
2119
+ "epoch": 0.6688417618270799,
2120
+ "grad_norm": 0.314453125,
2121
+ "learning_rate": 1.7020681945220796e-05,
2122
+ "loss": 0.5934,
2123
+ "step": 2460
2124
+ },
2125
+ {
2126
+ "epoch": 0.6715606307775965,
2127
+ "grad_norm": 0.28125,
2128
+ "learning_rate": 1.6880939072107325e-05,
2129
+ "loss": 0.5953,
2130
+ "step": 2470
2131
+ },
2132
+ {
2133
+ "epoch": 0.6742794997281131,
2134
+ "grad_norm": 0.28515625,
2135
+ "learning_rate": 1.6741196198993854e-05,
2136
+ "loss": 0.5883,
2137
+ "step": 2480
2138
+ },
2139
+ {
2140
+ "epoch": 0.6769983686786297,
2141
+ "grad_norm": 0.2890625,
2142
+ "learning_rate": 1.660145332588038e-05,
2143
+ "loss": 0.5918,
2144
+ "step": 2490
2145
+ },
2146
+ {
2147
+ "epoch": 0.6797172376291463,
2148
+ "grad_norm": 0.283203125,
2149
+ "learning_rate": 1.646171045276691e-05,
2150
+ "loss": 0.5941,
2151
+ "step": 2500
2152
+ },
2153
+ {
2154
+ "epoch": 0.6797172376291463,
2155
+ "eval_loss": 0.5862187743186951,
2156
+ "eval_runtime": 102.0266,
2157
+ "eval_samples_per_second": 4.901,
2158
+ "eval_steps_per_second": 0.157,
2159
+ "step": 2500
2160
+ },
2161
+ {
2162
+ "epoch": 0.6824361065796629,
2163
+ "grad_norm": 0.33203125,
2164
+ "learning_rate": 1.632196757965344e-05,
2165
+ "loss": 0.5867,
2166
+ "step": 2510
2167
+ },
2168
+ {
2169
+ "epoch": 0.6851549755301795,
2170
+ "grad_norm": 0.279296875,
2171
+ "learning_rate": 1.6182224706539968e-05,
2172
+ "loss": 0.5922,
2173
+ "step": 2520
2174
+ },
2175
+ {
2176
+ "epoch": 0.687873844480696,
2177
+ "grad_norm": 0.37109375,
2178
+ "learning_rate": 1.6042481833426497e-05,
2179
+ "loss": 0.5902,
2180
+ "step": 2530
2181
+ },
2182
+ {
2183
+ "epoch": 0.6905927134312126,
2184
+ "grad_norm": 0.279296875,
2185
+ "learning_rate": 1.5902738960313023e-05,
2186
+ "loss": 0.5813,
2187
+ "step": 2540
2188
+ },
2189
+ {
2190
+ "epoch": 0.6933115823817292,
2191
+ "grad_norm": 0.296875,
2192
+ "learning_rate": 1.5762996087199552e-05,
2193
+ "loss": 0.5953,
2194
+ "step": 2550
2195
+ },
2196
+ {
2197
+ "epoch": 0.6933115823817292,
2198
+ "eval_loss": 0.5857187509536743,
2199
+ "eval_runtime": 102.0545,
2200
+ "eval_samples_per_second": 4.899,
2201
+ "eval_steps_per_second": 0.157,
2202
+ "step": 2550
2203
+ },
2204
+ {
2205
+ "epoch": 0.6960304513322458,
2206
+ "grad_norm": 0.31640625,
2207
+ "learning_rate": 1.562325321408608e-05,
2208
+ "loss": 0.5898,
2209
+ "step": 2560
2210
+ },
2211
+ {
2212
+ "epoch": 0.6987493202827624,
2213
+ "grad_norm": 0.302734375,
2214
+ "learning_rate": 1.5483510340972614e-05,
2215
+ "loss": 0.5918,
2216
+ "step": 2570
2217
+ },
2218
+ {
2219
+ "epoch": 0.7014681892332789,
2220
+ "grad_norm": 0.375,
2221
+ "learning_rate": 1.534376746785914e-05,
2222
+ "loss": 0.5852,
2223
+ "step": 2580
2224
+ },
2225
+ {
2226
+ "epoch": 0.7041870581837956,
2227
+ "grad_norm": 0.322265625,
2228
+ "learning_rate": 1.5204024594745669e-05,
2229
+ "loss": 0.5863,
2230
+ "step": 2590
2231
+ },
2232
+ {
2233
+ "epoch": 0.7069059271343121,
2234
+ "grad_norm": 0.275390625,
2235
+ "learning_rate": 1.5064281721632198e-05,
2236
+ "loss": 0.591,
2237
+ "step": 2600
2238
+ },
2239
+ {
2240
+ "epoch": 0.7069059271343121,
2241
+ "eval_loss": 0.5842187404632568,
2242
+ "eval_runtime": 98.8852,
2243
+ "eval_samples_per_second": 5.056,
2244
+ "eval_steps_per_second": 0.162,
2245
+ "step": 2600
2246
+ },
2247
+ {
2248
+ "epoch": 0.7096247960848288,
2249
+ "grad_norm": 0.2890625,
2250
+ "learning_rate": 1.4924538848518726e-05,
2251
+ "loss": 0.5727,
2252
+ "step": 2610
2253
+ },
2254
+ {
2255
+ "epoch": 0.7123436650353453,
2256
+ "grad_norm": 0.4140625,
2257
+ "learning_rate": 1.4784795975405255e-05,
2258
+ "loss": 0.5859,
2259
+ "step": 2620
2260
+ },
2261
+ {
2262
+ "epoch": 0.7150625339858618,
2263
+ "grad_norm": 0.30859375,
2264
+ "learning_rate": 1.4645053102291784e-05,
2265
+ "loss": 0.5891,
2266
+ "step": 2630
2267
+ },
2268
+ {
2269
+ "epoch": 0.7177814029363785,
2270
+ "grad_norm": 0.255859375,
2271
+ "learning_rate": 1.4505310229178312e-05,
2272
+ "loss": 0.5914,
2273
+ "step": 2640
2274
+ },
2275
+ {
2276
+ "epoch": 0.720500271886895,
2277
+ "grad_norm": 0.296875,
2278
+ "learning_rate": 1.4365567356064841e-05,
2279
+ "loss": 0.5859,
2280
+ "step": 2650
2281
+ },
2282
+ {
2283
+ "epoch": 0.720500271886895,
2284
+ "eval_loss": 0.5845312476158142,
2285
+ "eval_runtime": 98.9031,
2286
+ "eval_samples_per_second": 5.055,
2287
+ "eval_steps_per_second": 0.162,
2288
+ "step": 2650
2289
+ },
2290
+ {
2291
+ "epoch": 0.7232191408374117,
2292
+ "grad_norm": 0.439453125,
2293
+ "learning_rate": 1.4225824482951369e-05,
2294
+ "loss": 0.5988,
2295
+ "step": 2660
2296
+ },
2297
+ {
2298
+ "epoch": 0.7259380097879282,
2299
+ "grad_norm": 0.287109375,
2300
+ "learning_rate": 1.4086081609837898e-05,
2301
+ "loss": 0.5926,
2302
+ "step": 2670
2303
+ },
2304
+ {
2305
+ "epoch": 0.7286568787384448,
2306
+ "grad_norm": 0.314453125,
2307
+ "learning_rate": 1.3946338736724427e-05,
2308
+ "loss": 0.5813,
2309
+ "step": 2680
2310
+ },
2311
+ {
2312
+ "epoch": 0.7313757476889614,
2313
+ "grad_norm": 0.30859375,
2314
+ "learning_rate": 1.3806595863610958e-05,
2315
+ "loss": 0.591,
2316
+ "step": 2690
2317
+ },
2318
+ {
2319
+ "epoch": 0.734094616639478,
2320
+ "grad_norm": 0.333984375,
2321
+ "learning_rate": 1.3666852990497486e-05,
2322
+ "loss": 0.5852,
2323
+ "step": 2700
2324
+ },
2325
+ {
2326
+ "epoch": 0.734094616639478,
2327
+ "eval_loss": 0.5846250057220459,
2328
+ "eval_runtime": 98.9658,
2329
+ "eval_samples_per_second": 5.052,
2330
+ "eval_steps_per_second": 0.162,
2331
+ "step": 2700
2332
+ },
2333
+ {
2334
+ "epoch": 0.7368134855899946,
2335
+ "grad_norm": 0.3046875,
2336
+ "learning_rate": 1.3527110117384015e-05,
2337
+ "loss": 0.5844,
2338
+ "step": 2710
2339
+ },
2340
+ {
2341
+ "epoch": 0.7395323545405111,
2342
+ "grad_norm": 0.2734375,
2343
+ "learning_rate": 1.3387367244270544e-05,
2344
+ "loss": 0.5887,
2345
+ "step": 2720
2346
+ },
2347
+ {
2348
+ "epoch": 0.7422512234910277,
2349
+ "grad_norm": 0.302734375,
2350
+ "learning_rate": 1.3247624371157072e-05,
2351
+ "loss": 0.5887,
2352
+ "step": 2730
2353
+ },
2354
+ {
2355
+ "epoch": 0.7449700924415443,
2356
+ "grad_norm": 0.275390625,
2357
+ "learning_rate": 1.3107881498043601e-05,
2358
+ "loss": 0.591,
2359
+ "step": 2740
2360
+ },
2361
+ {
2362
+ "epoch": 0.7476889613920609,
2363
+ "grad_norm": 0.26953125,
2364
+ "learning_rate": 1.2968138624930128e-05,
2365
+ "loss": 0.5938,
2366
+ "step": 2750
2367
+ },
2368
+ {
2369
+ "epoch": 0.7476889613920609,
2370
+ "eval_loss": 0.5854062438011169,
2371
+ "eval_runtime": 100.5592,
2372
+ "eval_samples_per_second": 4.972,
2373
+ "eval_steps_per_second": 0.159,
2374
+ "step": 2750
2375
+ },
2376
+ {
2377
+ "epoch": 0.7504078303425775,
2378
+ "grad_norm": 0.32421875,
2379
+ "learning_rate": 1.2828395751816658e-05,
2380
+ "loss": 0.5949,
2381
+ "step": 2760
2382
+ },
2383
+ {
2384
+ "epoch": 0.7531266992930941,
2385
+ "grad_norm": 0.34765625,
2386
+ "learning_rate": 1.2688652878703185e-05,
2387
+ "loss": 0.5957,
2388
+ "step": 2770
2389
+ },
2390
+ {
2391
+ "epoch": 0.7558455682436107,
2392
+ "grad_norm": 0.310546875,
2393
+ "learning_rate": 1.2548910005589715e-05,
2394
+ "loss": 0.5953,
2395
+ "step": 2780
2396
+ },
2397
+ {
2398
+ "epoch": 0.7585644371941273,
2399
+ "grad_norm": 0.3125,
2400
+ "learning_rate": 1.2409167132476244e-05,
2401
+ "loss": 0.5848,
2402
+ "step": 2790
2403
+ },
2404
+ {
2405
+ "epoch": 0.7612833061446438,
2406
+ "grad_norm": 0.443359375,
2407
+ "learning_rate": 1.2269424259362773e-05,
2408
+ "loss": 0.5953,
2409
+ "step": 2800
2410
+ },
2411
+ {
2412
+ "epoch": 0.7612833061446438,
2413
+ "eval_loss": 0.5840625166893005,
2414
+ "eval_runtime": 100.2413,
2415
+ "eval_samples_per_second": 4.988,
2416
+ "eval_steps_per_second": 0.16,
2417
+ "step": 2800
2418
+ },
2419
+ {
2420
+ "epoch": 0.7640021750951604,
2421
+ "grad_norm": 0.32421875,
2422
+ "learning_rate": 1.2129681386249302e-05,
2423
+ "loss": 0.5906,
2424
+ "step": 2810
2425
+ },
2426
+ {
2427
+ "epoch": 0.766721044045677,
2428
+ "grad_norm": 0.30078125,
2429
+ "learning_rate": 1.198993851313583e-05,
2430
+ "loss": 0.5887,
2431
+ "step": 2820
2432
+ },
2433
+ {
2434
+ "epoch": 0.7694399129961936,
2435
+ "grad_norm": 0.341796875,
2436
+ "learning_rate": 1.1850195640022359e-05,
2437
+ "loss": 0.5953,
2438
+ "step": 2830
2439
+ },
2440
+ {
2441
+ "epoch": 0.7721587819467102,
2442
+ "grad_norm": 0.3203125,
2443
+ "learning_rate": 1.1710452766908888e-05,
2444
+ "loss": 0.5906,
2445
+ "step": 2840
2446
+ },
2447
+ {
2448
+ "epoch": 0.7748776508972267,
2449
+ "grad_norm": 0.29296875,
2450
+ "learning_rate": 1.1570709893795418e-05,
2451
+ "loss": 0.5859,
2452
+ "step": 2850
2453
+ },
2454
+ {
2455
+ "epoch": 0.7748776508972267,
2456
+ "eval_loss": 0.581250011920929,
2457
+ "eval_runtime": 100.9043,
2458
+ "eval_samples_per_second": 4.955,
2459
+ "eval_steps_per_second": 0.159,
2460
+ "step": 2850
2461
+ },
2462
+ {
2463
+ "epoch": 0.7775965198477434,
2464
+ "grad_norm": 0.29296875,
2465
+ "learning_rate": 1.1430967020681945e-05,
2466
+ "loss": 0.5898,
2467
+ "step": 2860
2468
+ },
2469
+ {
2470
+ "epoch": 0.7803153887982599,
2471
+ "grad_norm": 0.30078125,
2472
+ "learning_rate": 1.1291224147568474e-05,
2473
+ "loss": 0.5914,
2474
+ "step": 2870
2475
+ },
2476
+ {
2477
+ "epoch": 0.7830342577487766,
2478
+ "grad_norm": 0.263671875,
2479
+ "learning_rate": 1.1151481274455004e-05,
2480
+ "loss": 0.5828,
2481
+ "step": 2880
2482
+ },
2483
+ {
2484
+ "epoch": 0.7857531266992931,
2485
+ "grad_norm": 0.306640625,
2486
+ "learning_rate": 1.1011738401341531e-05,
2487
+ "loss": 0.5891,
2488
+ "step": 2890
2489
+ },
2490
+ {
2491
+ "epoch": 0.7884719956498096,
2492
+ "grad_norm": 0.294921875,
2493
+ "learning_rate": 1.0871995528228062e-05,
2494
+ "loss": 0.5914,
2495
+ "step": 2900
2496
+ },
2497
+ {
2498
+ "epoch": 0.7884719956498096,
2499
+ "eval_loss": 0.582281231880188,
2500
+ "eval_runtime": 119.7467,
2501
+ "eval_samples_per_second": 4.175,
2502
+ "eval_steps_per_second": 0.134,
2503
+ "step": 2900
2504
+ },
2505
+ {
2506
+ "epoch": 0.7911908646003263,
2507
+ "grad_norm": 0.3359375,
2508
+ "learning_rate": 1.073225265511459e-05,
2509
+ "loss": 0.5914,
2510
+ "step": 2910
2511
+ },
2512
+ {
2513
+ "epoch": 0.7939097335508428,
2514
+ "grad_norm": 0.318359375,
2515
+ "learning_rate": 1.0592509782001119e-05,
2516
+ "loss": 0.5801,
2517
+ "step": 2920
2518
+ },
2519
+ {
2520
+ "epoch": 0.7966286025013595,
2521
+ "grad_norm": 0.283203125,
2522
+ "learning_rate": 1.0452766908887646e-05,
2523
+ "loss": 0.5891,
2524
+ "step": 2930
2525
+ },
2526
+ {
2527
+ "epoch": 0.799347471451876,
2528
+ "grad_norm": 0.376953125,
2529
+ "learning_rate": 1.0313024035774176e-05,
2530
+ "loss": 0.598,
2531
+ "step": 2940
2532
+ },
2533
+ {
2534
+ "epoch": 0.8020663404023926,
2535
+ "grad_norm": 0.32421875,
2536
+ "learning_rate": 1.0173281162660705e-05,
2537
+ "loss": 0.5848,
2538
+ "step": 2950
2539
+ },
2540
+ {
2541
+ "epoch": 0.8020663404023926,
2542
+ "eval_loss": 0.5870312452316284,
2543
+ "eval_runtime": 108.4768,
2544
+ "eval_samples_per_second": 4.609,
2545
+ "eval_steps_per_second": 0.147,
2546
+ "step": 2950
2547
+ },
2548
+ {
2549
+ "epoch": 0.8047852093529092,
2550
+ "grad_norm": 0.306640625,
2551
+ "learning_rate": 1.0033538289547234e-05,
2552
+ "loss": 0.5852,
2553
+ "step": 2960
2554
+ },
2555
+ {
2556
+ "epoch": 0.8075040783034257,
2557
+ "grad_norm": 0.5625,
2558
+ "learning_rate": 9.893795416433763e-06,
2559
+ "loss": 0.5875,
2560
+ "step": 2970
2561
+ },
2562
+ {
2563
+ "epoch": 0.8102229472539424,
2564
+ "grad_norm": 0.31640625,
2565
+ "learning_rate": 9.754052543320291e-06,
2566
+ "loss": 0.5949,
2567
+ "step": 2980
2568
+ },
2569
+ {
2570
+ "epoch": 0.8129418162044589,
2571
+ "grad_norm": 0.267578125,
2572
+ "learning_rate": 9.61430967020682e-06,
2573
+ "loss": 0.5895,
2574
+ "step": 2990
2575
+ },
2576
+ {
2577
+ "epoch": 0.8156606851549756,
2578
+ "grad_norm": 0.279296875,
2579
+ "learning_rate": 9.474566797093348e-06,
2580
+ "loss": 0.5836,
2581
+ "step": 3000
2582
+ },
2583
+ {
2584
+ "epoch": 0.8156606851549756,
2585
+ "eval_loss": 0.5854687690734863,
2586
+ "eval_runtime": 105.7926,
2587
+ "eval_samples_per_second": 4.726,
2588
+ "eval_steps_per_second": 0.151,
2589
+ "step": 3000
2590
+ },
2591
+ {
2592
+ "epoch": 0.8183795541054921,
2593
+ "grad_norm": 0.3046875,
2594
+ "learning_rate": 9.334823923979877e-06,
2595
+ "loss": 0.5867,
2596
+ "step": 3010
2597
+ },
2598
+ {
2599
+ "epoch": 0.8210984230560087,
2600
+ "grad_norm": 0.29296875,
2601
+ "learning_rate": 9.195081050866406e-06,
2602
+ "loss": 0.5926,
2603
+ "step": 3020
2604
+ },
2605
+ {
2606
+ "epoch": 0.8238172920065253,
2607
+ "grad_norm": 0.314453125,
2608
+ "learning_rate": 9.055338177752935e-06,
2609
+ "loss": 0.5828,
2610
+ "step": 3030
2611
+ },
2612
+ {
2613
+ "epoch": 0.8265361609570419,
2614
+ "grad_norm": 0.302734375,
2615
+ "learning_rate": 8.915595304639463e-06,
2616
+ "loss": 0.5926,
2617
+ "step": 3040
2618
+ },
2619
+ {
2620
+ "epoch": 0.8292550299075585,
2621
+ "grad_norm": 0.33203125,
2622
+ "learning_rate": 8.775852431525992e-06,
2623
+ "loss": 0.5949,
2624
+ "step": 3050
2625
+ },
2626
+ {
2627
+ "epoch": 0.8292550299075585,
2628
+ "eval_loss": 0.5823125243186951,
2629
+ "eval_runtime": 104.9839,
2630
+ "eval_samples_per_second": 4.763,
2631
+ "eval_steps_per_second": 0.152,
2632
+ "step": 3050
2633
+ },
2634
+ {
2635
+ "epoch": 0.831973898858075,
2636
+ "grad_norm": 0.384765625,
2637
+ "learning_rate": 8.636109558412521e-06,
2638
+ "loss": 0.5957,
2639
+ "step": 3060
2640
+ },
2641
+ {
2642
+ "epoch": 0.8346927678085916,
2643
+ "grad_norm": 0.5546875,
2644
+ "learning_rate": 8.496366685299049e-06,
2645
+ "loss": 0.5895,
2646
+ "step": 3070
2647
+ },
2648
+ {
2649
+ "epoch": 0.8374116367591082,
2650
+ "grad_norm": 0.30859375,
2651
+ "learning_rate": 8.35662381218558e-06,
2652
+ "loss": 0.5965,
2653
+ "step": 3080
2654
+ },
2655
+ {
2656
+ "epoch": 0.8401305057096248,
2657
+ "grad_norm": 0.318359375,
2658
+ "learning_rate": 8.216880939072108e-06,
2659
+ "loss": 0.5848,
2660
+ "step": 3090
2661
+ },
2662
+ {
2663
+ "epoch": 0.8428493746601414,
2664
+ "grad_norm": 0.337890625,
2665
+ "learning_rate": 8.077138065958637e-06,
2666
+ "loss": 0.5828,
2667
+ "step": 3100
2668
+ },
2669
+ {
2670
+ "epoch": 0.8428493746601414,
2671
+ "eval_loss": 0.5844687223434448,
2672
+ "eval_runtime": 103.8778,
2673
+ "eval_samples_per_second": 4.813,
2674
+ "eval_steps_per_second": 0.154,
2675
+ "step": 3100
2676
+ },
2677
+ {
2678
+ "epoch": 0.845568243610658,
2679
+ "grad_norm": 0.31640625,
2680
+ "learning_rate": 7.937395192845164e-06,
2681
+ "loss": 0.5898,
2682
+ "step": 3110
2683
+ },
2684
+ {
2685
+ "epoch": 0.8482871125611745,
2686
+ "grad_norm": 0.31640625,
2687
+ "learning_rate": 7.797652319731694e-06,
2688
+ "loss": 0.5863,
2689
+ "step": 3120
2690
+ },
2691
+ {
2692
+ "epoch": 0.8510059815116912,
2693
+ "grad_norm": 0.318359375,
2694
+ "learning_rate": 7.657909446618223e-06,
2695
+ "loss": 0.5844,
2696
+ "step": 3130
2697
+ },
2698
+ {
2699
+ "epoch": 0.8537248504622077,
2700
+ "grad_norm": 0.31640625,
2701
+ "learning_rate": 7.518166573504752e-06,
2702
+ "loss": 0.5832,
2703
+ "step": 3140
2704
+ },
2705
+ {
2706
+ "epoch": 0.8564437194127243,
2707
+ "grad_norm": 0.310546875,
2708
+ "learning_rate": 7.3784237003912805e-06,
2709
+ "loss": 0.591,
2710
+ "step": 3150
2711
+ },
2712
+ {
2713
+ "epoch": 0.8564437194127243,
2714
+ "eval_loss": 0.5841875076293945,
2715
+ "eval_runtime": 100.5886,
2716
+ "eval_samples_per_second": 4.971,
2717
+ "eval_steps_per_second": 0.159,
2718
+ "step": 3150
2719
+ },
2720
+ {
2721
+ "epoch": 0.8591625883632409,
2722
+ "grad_norm": 0.296875,
2723
+ "learning_rate": 7.238680827277809e-06,
2724
+ "loss": 0.5891,
2725
+ "step": 3160
2726
+ },
2727
+ {
2728
+ "epoch": 0.8618814573137574,
2729
+ "grad_norm": 0.318359375,
2730
+ "learning_rate": 7.098937954164338e-06,
2731
+ "loss": 0.5941,
2732
+ "step": 3170
2733
+ },
2734
+ {
2735
+ "epoch": 0.8646003262642741,
2736
+ "grad_norm": 0.30859375,
2737
+ "learning_rate": 6.9591950810508665e-06,
2738
+ "loss": 0.5848,
2739
+ "step": 3180
2740
+ },
2741
+ {
2742
+ "epoch": 0.8673191952147906,
2743
+ "grad_norm": 0.28515625,
2744
+ "learning_rate": 6.819452207937395e-06,
2745
+ "loss": 0.5918,
2746
+ "step": 3190
2747
+ },
2748
+ {
2749
+ "epoch": 0.8700380641653073,
2750
+ "grad_norm": 0.306640625,
2751
+ "learning_rate": 6.679709334823925e-06,
2752
+ "loss": 0.5867,
2753
+ "step": 3200
2754
+ },
2755
+ {
2756
+ "epoch": 0.8700380641653073,
2757
+ "eval_loss": 0.5841562747955322,
2758
+ "eval_runtime": 100.4222,
2759
+ "eval_samples_per_second": 4.979,
2760
+ "eval_steps_per_second": 0.159,
2761
+ "step": 3200
2762
+ },
2763
+ {
2764
+ "epoch": 0.8727569331158238,
2765
+ "grad_norm": 0.359375,
2766
+ "learning_rate": 6.539966461710453e-06,
2767
+ "loss": 0.5934,
2768
+ "step": 3210
2769
+ },
2770
+ {
2771
+ "epoch": 0.8754758020663403,
2772
+ "grad_norm": 0.3046875,
2773
+ "learning_rate": 6.400223588596982e-06,
2774
+ "loss": 0.5863,
2775
+ "step": 3220
2776
+ },
2777
+ {
2778
+ "epoch": 0.878194671016857,
2779
+ "grad_norm": 0.302734375,
2780
+ "learning_rate": 6.26048071548351e-06,
2781
+ "loss": 0.5793,
2782
+ "step": 3230
2783
+ },
2784
+ {
2785
+ "epoch": 0.8809135399673735,
2786
+ "grad_norm": 0.328125,
2787
+ "learning_rate": 6.1207378423700394e-06,
2788
+ "loss": 0.5801,
2789
+ "step": 3240
2790
+ },
2791
+ {
2792
+ "epoch": 0.8836324089178902,
2793
+ "grad_norm": 0.306640625,
2794
+ "learning_rate": 5.980994969256568e-06,
2795
+ "loss": 0.5891,
2796
+ "step": 3250
2797
+ },
2798
+ {
2799
+ "epoch": 0.8836324089178902,
2800
+ "eval_loss": 0.5814999938011169,
2801
+ "eval_runtime": 98.5014,
2802
+ "eval_samples_per_second": 5.076,
2803
+ "eval_steps_per_second": 0.162,
2804
+ "step": 3250
2805
+ },
2806
+ {
2807
+ "epoch": 0.8863512778684067,
2808
+ "grad_norm": 0.359375,
2809
+ "learning_rate": 5.841252096143097e-06,
2810
+ "loss": 0.5895,
2811
+ "step": 3260
2812
+ },
2813
+ {
2814
+ "epoch": 0.8890701468189234,
2815
+ "grad_norm": 0.3046875,
2816
+ "learning_rate": 5.701509223029626e-06,
2817
+ "loss": 0.5949,
2818
+ "step": 3270
2819
+ },
2820
+ {
2821
+ "epoch": 0.8917890157694399,
2822
+ "grad_norm": 0.34375,
2823
+ "learning_rate": 5.561766349916155e-06,
2824
+ "loss": 0.5891,
2825
+ "step": 3280
2826
+ },
2827
+ {
2828
+ "epoch": 0.8945078847199565,
2829
+ "grad_norm": 0.296875,
2830
+ "learning_rate": 5.422023476802683e-06,
2831
+ "loss": 0.5848,
2832
+ "step": 3290
2833
+ },
2834
+ {
2835
+ "epoch": 0.8972267536704731,
2836
+ "grad_norm": 0.34765625,
2837
+ "learning_rate": 5.282280603689212e-06,
2838
+ "loss": 0.5953,
2839
+ "step": 3300
2840
+ },
2841
+ {
2842
+ "epoch": 0.8972267536704731,
2843
+ "eval_loss": 0.5840625166893005,
2844
+ "eval_runtime": 98.7779,
2845
+ "eval_samples_per_second": 5.062,
2846
+ "eval_steps_per_second": 0.162,
2847
+ "step": 3300
2848
+ },
2849
+ {
2850
+ "epoch": 0.8999456226209896,
2851
+ "grad_norm": 0.3359375,
2852
+ "learning_rate": 5.142537730575741e-06,
2853
+ "loss": 0.5922,
2854
+ "step": 3310
2855
+ },
2856
+ {
2857
+ "epoch": 0.9026644915715063,
2858
+ "grad_norm": 0.328125,
2859
+ "learning_rate": 5.002794857462269e-06,
2860
+ "loss": 0.5871,
2861
+ "step": 3320
2862
+ },
2863
+ {
2864
+ "epoch": 0.9053833605220228,
2865
+ "grad_norm": 0.345703125,
2866
+ "learning_rate": 4.863051984348798e-06,
2867
+ "loss": 0.5891,
2868
+ "step": 3330
2869
+ },
2870
+ {
2871
+ "epoch": 0.9081022294725394,
2872
+ "grad_norm": 0.3046875,
2873
+ "learning_rate": 4.723309111235328e-06,
2874
+ "loss": 0.5988,
2875
+ "step": 3340
2876
+ },
2877
+ {
2878
+ "epoch": 0.910821098423056,
2879
+ "grad_norm": 0.30078125,
2880
+ "learning_rate": 4.583566238121856e-06,
2881
+ "loss": 0.5797,
2882
+ "step": 3350
2883
+ },
2884
+ {
2885
+ "epoch": 0.910821098423056,
2886
+ "eval_loss": 0.5815625190734863,
2887
+ "eval_runtime": 97.4891,
2888
+ "eval_samples_per_second": 5.129,
2889
+ "eval_steps_per_second": 0.164,
2890
+ "step": 3350
2891
+ },
2892
+ {
2893
+ "epoch": 0.9135399673735726,
2894
+ "grad_norm": 0.32421875,
2895
+ "learning_rate": 4.443823365008385e-06,
2896
+ "loss": 0.5785,
2897
+ "step": 3360
2898
+ },
2899
+ {
2900
+ "epoch": 0.9162588363240892,
2901
+ "grad_norm": 0.302734375,
2902
+ "learning_rate": 4.304080491894914e-06,
2903
+ "loss": 0.5844,
2904
+ "step": 3370
2905
+ },
2906
+ {
2907
+ "epoch": 0.9189777052746058,
2908
+ "grad_norm": 0.3125,
2909
+ "learning_rate": 4.164337618781442e-06,
2910
+ "loss": 0.5867,
2911
+ "step": 3380
2912
+ },
2913
+ {
2914
+ "epoch": 0.9216965742251223,
2915
+ "grad_norm": 0.2890625,
2916
+ "learning_rate": 4.024594745667971e-06,
2917
+ "loss": 0.584,
2918
+ "step": 3390
2919
+ },
2920
+ {
2921
+ "epoch": 0.924415443175639,
2922
+ "grad_norm": 0.314453125,
2923
+ "learning_rate": 3.8848518725545e-06,
2924
+ "loss": 0.5859,
2925
+ "step": 3400
2926
+ },
2927
+ {
2928
+ "epoch": 0.924415443175639,
2929
+ "eval_loss": 0.5829687714576721,
2930
+ "eval_runtime": 108.7539,
2931
+ "eval_samples_per_second": 4.598,
2932
+ "eval_steps_per_second": 0.147,
2933
+ "step": 3400
2934
+ },
2935
+ {
2936
+ "epoch": 0.9271343121261555,
2937
+ "grad_norm": 0.30078125,
2938
+ "learning_rate": 3.7451089994410285e-06,
2939
+ "loss": 0.591,
2940
+ "step": 3410
2941
+ },
2942
+ {
2943
+ "epoch": 0.9298531810766721,
2944
+ "grad_norm": 0.322265625,
2945
+ "learning_rate": 3.6053661263275578e-06,
2946
+ "loss": 0.5863,
2947
+ "step": 3420
2948
+ },
2949
+ {
2950
+ "epoch": 0.9325720500271887,
2951
+ "grad_norm": 0.2890625,
2952
+ "learning_rate": 3.465623253214086e-06,
2953
+ "loss": 0.5801,
2954
+ "step": 3430
2955
+ },
2956
+ {
2957
+ "epoch": 0.9352909189777052,
2958
+ "grad_norm": 0.40625,
2959
+ "learning_rate": 3.325880380100615e-06,
2960
+ "loss": 0.5961,
2961
+ "step": 3440
2962
+ },
2963
+ {
2964
+ "epoch": 0.9380097879282219,
2965
+ "grad_norm": 0.296875,
2966
+ "learning_rate": 3.1861375069871442e-06,
2967
+ "loss": 0.5891,
2968
+ "step": 3450
2969
+ },
2970
+ {
2971
+ "epoch": 0.9380097879282219,
2972
+ "eval_loss": 0.5825625061988831,
2973
+ "eval_runtime": 98.4443,
2974
+ "eval_samples_per_second": 5.079,
2975
+ "eval_steps_per_second": 0.163,
2976
+ "step": 3450
2977
+ },
2978
+ {
2979
+ "epoch": 0.9407286568787384,
2980
+ "grad_norm": 0.29296875,
2981
+ "learning_rate": 3.0463946338736726e-06,
2982
+ "loss": 0.5895,
2983
+ "step": 3460
2984
+ },
2985
+ {
2986
+ "epoch": 0.9434475258292551,
2987
+ "grad_norm": 0.291015625,
2988
+ "learning_rate": 2.9066517607602015e-06,
2989
+ "loss": 0.5891,
2990
+ "step": 3470
2991
+ },
2992
+ {
2993
+ "epoch": 0.9461663947797716,
2994
+ "grad_norm": 0.36328125,
2995
+ "learning_rate": 2.7669088876467303e-06,
2996
+ "loss": 0.5941,
2997
+ "step": 3480
2998
+ },
2999
+ {
3000
+ "epoch": 0.9488852637302883,
3001
+ "grad_norm": 0.3125,
3002
+ "learning_rate": 2.627166014533259e-06,
3003
+ "loss": 0.5875,
3004
+ "step": 3490
3005
+ },
3006
+ {
3007
+ "epoch": 0.9516041326808048,
3008
+ "grad_norm": 0.294921875,
3009
+ "learning_rate": 2.487423141419788e-06,
3010
+ "loss": 0.5875,
3011
+ "step": 3500
3012
+ },
3013
+ {
3014
+ "epoch": 0.9516041326808048,
3015
+ "eval_loss": 0.585687518119812,
3016
+ "eval_runtime": 99.0514,
3017
+ "eval_samples_per_second": 5.048,
3018
+ "eval_steps_per_second": 0.162,
3019
+ "step": 3500
3020
+ }
3021
+ ],
3022
+ "logging_steps": 10,
3023
+ "max_steps": 3678,
3024
+ "num_input_tokens_seen": 0,
3025
+ "num_train_epochs": 1,
3026
+ "save_steps": 500,
3027
+ "stateful_callbacks": {
3028
+ "TrainerControl": {
3029
+ "args": {
3030
+ "should_epoch_stop": false,
3031
+ "should_evaluate": false,
3032
+ "should_log": false,
3033
+ "should_save": true,
3034
+ "should_training_stop": false
3035
+ },
3036
+ "attributes": {}
3037
+ }
3038
+ },
3039
+ "total_flos": 1.68583631387136e+18,
3040
+ "train_batch_size": 32,
3041
+ "trial_name": null,
3042
+ "trial_params": null
3043
+ }
checkpoint-3500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a2cb32125f0d1a068d014dfe59cb834161e37ad4038e1e0aa9b327afed437892
3
+ size 5048
git_hash.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ b20318f7a457b3ef014beca2b279f716f99df188
preprocessor_config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_valid_processor_keys": [
3
+ "images",
4
+ "do_resize",
5
+ "size",
6
+ "resample",
7
+ "do_rescale",
8
+ "rescale_factor",
9
+ "do_normalize",
10
+ "image_mean",
11
+ "image_std",
12
+ "return_tensors",
13
+ "data_format",
14
+ "input_data_format",
15
+ "do_convert_rgb"
16
+ ],
17
+ "do_convert_rgb": null,
18
+ "do_normalize": true,
19
+ "do_rescale": true,
20
+ "do_resize": true,
21
+ "image_mean": [
22
+ 0.5,
23
+ 0.5,
24
+ 0.5
25
+ ],
26
+ "image_processor_type": "SiglipImageProcessor",
27
+ "image_seq_length": 1024,
28
+ "image_std": [
29
+ 0.5,
30
+ 0.5,
31
+ 0.5
32
+ ],
33
+ "processor_class": "PaliGemmaProcessor",
34
+ "resample": 3,
35
+ "rescale_factor": 0.00392156862745098,
36
+ "size": {
37
+ "height": 448,
38
+ "width": 448
39
+ }
40
+ }
results.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"validation_set": {"ndcg_at_1": 0.664, "ndcg_at_3": 0.71028, "ndcg_at_5": 0.72836, "ndcg_at_10": 0.74133, "ndcg_at_20": 0.75388, "ndcg_at_100": 0.76933, "ndcg_at_1000": 0.77491, "map_at_1": 0.664, "map_at_3": 0.69867, "map_at_5": 0.70867, "map_at_10": 0.71405, "map_at_20": 0.71744, "map_at_100": 0.71968, "map_at_1000": 0.71996, "recall_at_1": 0.664, "recall_at_3": 0.744, "recall_at_5": 0.788, "recall_at_10": 0.828, "recall_at_20": 0.878, "recall_at_100": 0.96, "recall_at_1000": 1.0, "precision_at_1": 0.664, "precision_at_3": 0.248, "precision_at_5": 0.1576, "precision_at_10": 0.0828, "precision_at_20": 0.0439, "precision_at_100": 0.0096, "precision_at_1000": 0.001, "mrr_at_1": 0.66, "mrr_at_3": 0.6969999999999998, "mrr_at_5": 0.7063999999999997, "mrr_at_10": 0.712376984126984, "mrr_at_20": 0.7157380512418438, "mrr_at_100": 0.7178831847626113, "mrr_at_1000": 0.7181677696156411, "naucs_at_1_max": 0.44018051894177584, "naucs_at_1_std": -0.7856086659797401, "naucs_at_1_diff1": 0.8849576469643642, "naucs_at_3_max": 0.39376468076019605, "naucs_at_3_std": -0.9284235943428755, "naucs_at_3_diff1": 0.8724231672662163, "naucs_at_5_max": 0.35038332649105774, "naucs_at_5_std": -0.9648370743071024, "naucs_at_5_diff1": 0.8496911429144268, "naucs_at_10_max": 0.3239955895527529, "naucs_at_10_std": -1.0699009946477418, "naucs_at_10_diff1": 0.8547481221142577, "naucs_at_20_max": 0.2038296938439574, "naucs_at_20_std": -1.2576774153093753, "naucs_at_20_diff1": 0.8449624555188024, "naucs_at_100_max": 0.103057889822593, "naucs_at_100_std": -1.3548552754435126, "naucs_at_100_diff1": 0.830182072829127, "naucs_at_1000_max": 1.0, "naucs_at_1000_std": 1.0, "naucs_at_1000_diff1": 1.0}, "syntheticDocQA_energy": {"ndcg_at_1": 0.71, "ndcg_at_3": 0.76917, "ndcg_at_5": 0.79457, "ndcg_at_10": 0.80417, "ndcg_at_20": 0.81686, "ndcg_at_100": 0.8265, "ndcg_at_1000": 0.8265, "map_at_1": 0.71, "map_at_3": 0.755, "map_at_5": 0.7695, "map_at_10": 0.77342, "map_at_20": 0.77692, "map_at_100": 0.77835, "map_at_1000": 0.77835, "recall_at_1": 0.71, "recall_at_3": 0.81, "recall_at_5": 0.87, "recall_at_10": 0.9, "recall_at_20": 0.95, "recall_at_100": 1.0, "recall_at_1000": 1.0, "precision_at_1": 0.71, "precision_at_3": 0.27, "precision_at_5": 0.174, "precision_at_10": 0.09, "precision_at_20": 0.0475, "precision_at_100": 0.01, "precision_at_1000": 0.001, "mrr_at_1": 0.71, "mrr_at_3": 0.7583333333333334, "mrr_at_5": 0.7708333333333335, "mrr_at_10": 0.7748611111111113, "mrr_at_20": 0.7783654401154404, "mrr_at_100": 0.7797901809229121, "mrr_at_1000": 0.7797901809229121, "naucs_at_1_max": 0.49871208049960203, "naucs_at_1_std": -0.3303003910086542, "naucs_at_1_diff1": 0.7304264032763191, "naucs_at_3_max": 0.5121274032165114, "naucs_at_3_std": -0.2902194981402899, "naucs_at_3_diff1": 0.70967572947771, "naucs_at_5_max": 0.4480115529882247, "naucs_at_5_std": -0.3741761090128093, "naucs_at_5_diff1": 0.6262312078797302, "naucs_at_10_max": 0.5887955182072843, "naucs_at_10_std": -0.3164332399626486, "naucs_at_10_diff1": 0.5903361344537817, "naucs_at_20_max": 0.5164332399626562, "naucs_at_20_std": -0.5164332399626398, "naucs_at_20_diff1": 0.43650793650794084, "naucs_at_100_max": NaN, "naucs_at_100_std": NaN, "naucs_at_100_diff1": NaN, "naucs_at_1000_max": NaN, "naucs_at_1000_std": NaN, "naucs_at_1000_diff1": NaN}, "syntheticDocQA_healthcare_industry": {"ndcg_at_1": 0.72, "ndcg_at_3": 0.83333, "ndcg_at_5": 0.84625, "ndcg_at_10": 0.84926, "ndcg_at_20": 0.85738, "ndcg_at_100": 0.86134, "ndcg_at_1000": 0.86134, "map_at_1": 0.72, "map_at_3": 0.80667, "map_at_5": 0.81417, "map_at_10": 0.81528, "map_at_20": 0.81779, "map_at_100": 0.81842, "map_at_1000": 0.81842, "recall_at_1": 0.72, "recall_at_3": 0.91, "recall_at_5": 0.94, "recall_at_10": 0.95, "recall_at_20": 0.98, "recall_at_100": 1.0, "recall_at_1000": 1.0, "precision_at_1": 0.72, "precision_at_3": 0.30333, "precision_at_5": 0.188, "precision_at_10": 0.095, "precision_at_20": 0.049, "precision_at_100": 0.01, "precision_at_1000": 0.001, "mrr_at_1": 0.73, "mrr_at_3": 0.8116666666666664, "mrr_at_5": 0.8191666666666664, "mrr_at_10": 0.8212777777777777, "mrr_at_20": 0.8228803418803418, "mrr_at_100": 0.8235071225071223, "mrr_at_1000": 0.8235071225071223, "naucs_at_1_max": 0.4841842185592191, "naucs_at_1_std": -0.29973672161172155, "naucs_at_1_diff1": 0.843692765567766, "naucs_at_3_max": 0.6206037970743837, "naucs_at_3_std": -0.4326174914410215, "naucs_at_3_diff1": 0.7785558667911598, "naucs_at_5_max": 0.8358232181761603, "naucs_at_5_std": -0.3345004668534105, "naucs_at_5_diff1": 0.8821195144724553, "naucs_at_10_max": 0.892156862745101, "naucs_at_10_std": -0.5122315592903719, "naucs_at_10_diff1": 0.8585434173669475, "naucs_at_20_max": 0.7957516339869297, "naucs_at_20_std": -1.1517273576096962, "naucs_at_20_diff1": 0.7770774976657324, "naucs_at_100_max": NaN, "naucs_at_100_std": NaN, "naucs_at_100_diff1": NaN, "naucs_at_1000_max": NaN, "naucs_at_1000_std": NaN, "naucs_at_1000_diff1": NaN}, "syntheticDocQA_artificial_intelligence_test": {"ndcg_at_1": 0.7, "ndcg_at_3": 0.78178, "ndcg_at_5": 0.80288, "ndcg_at_10": 0.82014, "ndcg_at_20": 0.82264, "ndcg_at_100": 0.83196, "ndcg_at_1000": 0.83196, "map_at_1": 0.7, "map_at_3": 0.76167, "map_at_5": 0.77367, "map_at_10": 0.78144, "map_at_20": 0.78211, "map_at_100": 0.78339, "map_at_1000": 0.78339, "recall_at_1": 0.7, "recall_at_3": 0.84, "recall_at_5": 0.89, "recall_at_10": 0.94, "recall_at_20": 0.95, "recall_at_100": 1.0, "recall_at_1000": 1.0, "precision_at_1": 0.7, "precision_at_3": 0.28, "precision_at_5": 0.178, "precision_at_10": 0.094, "precision_at_20": 0.0475, "precision_at_100": 0.01, "precision_at_1000": 0.001, "mrr_at_1": 0.7, "mrr_at_3": 0.7616666666666667, "mrr_at_5": 0.7776666666666666, "mrr_at_10": 0.7821111111111112, "mrr_at_20": 0.7827777777777779, "mrr_at_100": 0.7840590679145849, "mrr_at_1000": 0.7840590679145849, "naucs_at_1_max": 0.590254008286795, "naucs_at_1_std": -0.15807962529273997, "naucs_at_1_diff1": 0.804395604395605, "naucs_at_3_max": 0.4802098674521348, "naucs_at_3_std": -0.2761413843888074, "naucs_at_3_diff1": 0.7229381443298973, "naucs_at_5_max": 0.4887896228846315, "naucs_at_5_std": -0.18159951894167012, "naucs_at_5_diff1": 0.7367064685164499, "naucs_at_10_max": 0.8883442265795238, "naucs_at_10_std": 0.23840647370058923, "naucs_at_10_diff1": 0.7833022097727997, "naucs_at_20_max": 0.8660130718954264, "naucs_at_20_std": 0.08608776844071178, "naucs_at_20_diff1": 0.8291316526610654, "naucs_at_100_max": NaN, "naucs_at_100_std": NaN, "naucs_at_100_diff1": NaN, "naucs_at_1000_max": NaN, "naucs_at_1000_std": NaN, "naucs_at_1000_diff1": NaN}, "syntheticDocQA_government_reports": {"ndcg_at_1": 0.68, "ndcg_at_3": 0.80333, "ndcg_at_5": 0.8399, "ndcg_at_10": 0.84323, "ndcg_at_20": 0.84323, "ndcg_at_100": 0.84323, "ndcg_at_1000": 0.84451, "map_at_1": 0.68, "map_at_3": 0.77333, "map_at_5": 0.79333, "map_at_10": 0.79476, "map_at_20": 0.79476, "map_at_100": 0.79476, "map_at_1000": 0.79481, "recall_at_1": 0.68, "recall_at_3": 0.89, "recall_at_5": 0.98, "recall_at_10": 0.99, "recall_at_20": 0.99, "recall_at_100": 0.99, "recall_at_1000": 1.0, "precision_at_1": 0.68, "precision_at_3": 0.29667, "precision_at_5": 0.196, "precision_at_10": 0.099, "precision_at_20": 0.0495, "precision_at_100": 0.0099, "precision_at_1000": 0.001, "mrr_at_1": 0.68, "mrr_at_3": 0.7733333333333334, "mrr_at_5": 0.7938333333333333, "mrr_at_10": 0.7952619047619046, "mrr_at_20": 0.7952619047619046, "mrr_at_100": 0.7952619047619046, "mrr_at_1000": 0.7953061525495153, "naucs_at_1_max": 0.3230944647916522, "naucs_at_1_std": -0.21867873678391347, "naucs_at_1_diff1": 0.7658247529541842, "naucs_at_3_max": 0.2032471437161745, "naucs_at_3_std": -0.5294218709732814, "naucs_at_3_diff1": 0.5349626320762807, "naucs_at_5_max": 0.540149393090577, "naucs_at_5_std": -0.690943043884218, "naucs_at_5_diff1": 0.4960317460317504, "naucs_at_10_max": 0.7222222222222276, "naucs_at_10_std": -1.7399626517273863, "naucs_at_10_diff1": 0.8692810457516413, "naucs_at_20_max": 0.7222222222222276, "naucs_at_20_std": -1.7399626517273863, "naucs_at_20_diff1": 0.8692810457516413, "naucs_at_100_max": 0.7222222222222041, "naucs_at_100_std": -1.7399626517273008, "naucs_at_100_diff1": 0.8692810457516374, "naucs_at_1000_max": NaN, "naucs_at_1000_std": NaN, "naucs_at_1000_diff1": NaN}, "infovqa_subsampled": {"ndcg_at_1": 0.62, "ndcg_at_3": 0.70578, "ndcg_at_5": 0.71844, "ndcg_at_10": 0.74042, "ndcg_at_20": 0.75054, "ndcg_at_100": 0.7593, "ndcg_at_1000": 0.76485, "map_at_1": 0.62, "map_at_3": 0.68433, "map_at_5": 0.69153, "map_at_10": 0.70021, "map_at_20": 0.70299, "map_at_100": 0.70412, "map_at_1000": 0.70432, "recall_at_1": 0.62, "recall_at_3": 0.768, "recall_at_5": 0.798, "recall_at_10": 0.868, "recall_at_20": 0.908, "recall_at_100": 0.956, "recall_at_1000": 1.0, "precision_at_1": 0.62, "precision_at_3": 0.256, "precision_at_5": 0.1596, "precision_at_10": 0.0868, "precision_at_20": 0.0454, "precision_at_100": 0.00956, "precision_at_1000": 0.001, "mrr_at_1": 0.622, "mrr_at_3": 0.6849999999999998, "mrr_at_5": 0.6930999999999998, "mrr_at_10": 0.7014198412698407, "mrr_at_20": 0.7042074367153308, "mrr_at_100": 0.7053358112690798, "mrr_at_1000": 0.7055319275265294, "naucs_at_1_max": 0.4713318770575025, "naucs_at_1_std": -0.027317478386610635, "naucs_at_1_diff1": 0.7819308986862752, "naucs_at_3_max": 0.5651872399445215, "naucs_at_3_std": 0.05637647142501536, "naucs_at_3_diff1": 0.665848714392404, "naucs_at_5_max": 0.6023633560846012, "naucs_at_5_std": 0.12076016000237819, "naucs_at_5_diff1": 0.622234779203656, "naucs_at_10_max": 0.6885647079449483, "naucs_at_10_std": 0.2404158022382579, "naucs_at_10_diff1": 0.5789162259299289, "naucs_at_20_max": 0.8001359964275557, "naucs_at_20_std": 0.45553728737871985, "naucs_at_20_diff1": 0.6461657126618755, "naucs_at_100_max": 0.8855148119854007, "naucs_at_100_std": 0.7027629233511608, "naucs_at_100_diff1": 0.6173711909006029, "naucs_at_1000_max": 1.0, "naucs_at_1000_std": 1.0, "naucs_at_1000_diff1": 1.0}, "docvqa_subsampled": {"ndcg_at_1": 0.18, "ndcg_at_3": 0.22371, "ndcg_at_5": 0.24806, "ndcg_at_10": 0.26793, "ndcg_at_20": 0.29433, "ndcg_at_100": 0.336, "ndcg_at_1000": 0.37198, "map_at_1": 0.18, "map_at_3": 0.212, "map_at_5": 0.2253, "map_at_10": 0.23338, "map_at_20": 0.24068, "map_at_100": 0.24599, "map_at_1000": 0.24728, "recall_at_1": 0.18, "recall_at_3": 0.258, "recall_at_5": 0.318, "recall_at_10": 0.38, "recall_at_20": 0.484, "recall_at_100": 0.716, "recall_at_1000": 1.0, "precision_at_1": 0.18, "precision_at_3": 0.086, "precision_at_5": 0.0636, "precision_at_10": 0.038, "precision_at_20": 0.0242, "precision_at_100": 0.00716, "precision_at_1000": 0.001, "mrr_at_1": 0.182, "mrr_at_3": 0.21266666666666653, "mrr_at_5": 0.22596666666666662, "mrr_at_10": 0.23399047619047622, "mrr_at_20": 0.24130295750001637, "mrr_at_100": 0.24662778080241265, "mrr_at_1000": 0.24791083548342926, "naucs_at_1_max": 0.3831175665719397, "naucs_at_1_std": -0.17559555134547236, "naucs_at_1_diff1": 0.633649089569519, "naucs_at_3_max": 0.2951265292150055, "naucs_at_3_std": -0.23959756660081263, "naucs_at_3_diff1": 0.5870561033701477, "naucs_at_5_max": 0.2599095226233618, "naucs_at_5_std": -0.18332248569048581, "naucs_at_5_diff1": 0.5445275650657687, "naucs_at_10_max": 0.23812803784472233, "naucs_at_10_std": -0.14582174511995405, "naucs_at_10_diff1": 0.5154783874405425, "naucs_at_20_max": 0.24329409519973313, "naucs_at_20_std": -0.1204545334089475, "naucs_at_20_diff1": 0.4988949368497487, "naucs_at_100_max": 0.26714534607465296, "naucs_at_100_std": 0.05045743549518379, "naucs_at_100_diff1": 0.4056732560507438, "naucs_at_1000_max": 1.0, "naucs_at_1000_std": 1.0, "naucs_at_1000_diff1": 1.0}, "arxivqa_subsampled": {"ndcg_at_1": 0.528, "ndcg_at_3": 0.59217, "ndcg_at_5": 0.60982, "ndcg_at_10": 0.6297, "ndcg_at_20": 0.64174, "ndcg_at_100": 0.65893, "ndcg_at_1000": 0.67322, "map_at_1": 0.528, "map_at_3": 0.57633, "map_at_5": 0.58633, "map_at_10": 0.59444, "map_at_20": 0.59769, "map_at_100": 0.59972, "map_at_1000": 0.60025, "recall_at_1": 0.528, "recall_at_3": 0.638, "recall_at_5": 0.68, "recall_at_10": 0.742, "recall_at_20": 0.79, "recall_at_100": 0.888, "recall_at_1000": 1.0, "precision_at_1": 0.528, "precision_at_3": 0.21267, "precision_at_5": 0.136, "precision_at_10": 0.0742, "precision_at_20": 0.0395, "precision_at_100": 0.00888, "precision_at_1000": 0.001, "mrr_at_1": 0.526, "mrr_at_3": 0.5753333333333335, "mrr_at_5": 0.5857333333333333, "mrr_at_10": 0.593568253968254, "mrr_at_20": 0.5968407763804824, "mrr_at_100": 0.598859750191098, "mrr_at_1000": 0.5993942947096509, "naucs_at_1_max": 0.7300555262926871, "naucs_at_1_std": -0.27287167500658427, "naucs_at_1_diff1": 0.8437609735756297, "naucs_at_3_max": 0.7465316990459587, "naucs_at_3_std": -0.25011849961984584, "naucs_at_3_diff1": 0.777427492548044, "naucs_at_5_max": 0.7391714463409579, "naucs_at_5_std": -0.2646534448206762, "naucs_at_5_diff1": 0.7407090042153271, "naucs_at_10_max": 0.7359005643446794, "naucs_at_10_std": -0.2943977671047861, "naucs_at_10_diff1": 0.7142612307855533, "naucs_at_20_max": 0.7230576862961985, "naucs_at_20_std": -0.27414336210835094, "naucs_at_20_diff1": 0.7045038112871814, "naucs_at_100_max": 0.721464082792209, "naucs_at_100_std": -0.3960447104978358, "naucs_at_100_diff1": 0.6451484713203481, "naucs_at_1000_max": 1.0, "naucs_at_1000_std": 1.0, "naucs_at_1000_diff1": 1.0}, "tabfquad_subsampled": {"ndcg_at_1": 0.59643, "ndcg_at_3": 0.68592, "ndcg_at_5": 0.7036, "ndcg_at_10": 0.73013, "ndcg_at_20": 0.73738, "ndcg_at_100": 0.75631, "ndcg_at_1000": 0.75631, "map_at_1": 0.59643, "map_at_3": 0.66488, "map_at_5": 0.6747, "map_at_10": 0.68563, "map_at_20": 0.68762, "map_at_100": 0.69033, "map_at_1000": 0.69033, "recall_at_1": 0.59643, "recall_at_3": 0.74643, "recall_at_5": 0.78929, "recall_at_10": 0.87143, "recall_at_20": 0.9, "recall_at_100": 1.0, "recall_at_1000": 1.0, "precision_at_1": 0.59643, "precision_at_3": 0.24881, "precision_at_5": 0.15786, "precision_at_10": 0.08714, "precision_at_20": 0.045, "precision_at_100": 0.01, "precision_at_1000": 0.001, "mrr_at_1": 0.5928571428571429, "mrr_at_3": 0.6619047619047619, "mrr_at_5": 0.6717261904761905, "mrr_at_10": 0.6828032879818592, "mrr_at_20": 0.6847972148465569, "mrr_at_100": 0.6875127980343567, "mrr_at_1000": 0.6875127980343567, "naucs_at_1_max": 0.5306576009206518, "naucs_at_1_std": 0.05700228579271614, "naucs_at_1_diff1": 0.6933553642223288, "naucs_at_3_max": 0.40943857756231683, "naucs_at_3_std": -0.049136472135799185, "naucs_at_3_diff1": 0.5201859337703321, "naucs_at_5_max": 0.3807182684813242, "naucs_at_5_std": -0.06477777872945445, "naucs_at_5_diff1": 0.53999468964395, "naucs_at_10_max": 0.3966235708943276, "naucs_at_10_std": 0.19033817715567913, "naucs_at_10_diff1": 0.5489101399722216, "naucs_at_20_max": 0.42121848739495904, "naucs_at_20_std": 0.1893590769641199, "naucs_at_20_diff1": 0.5727457649726578, "naucs_at_100_max": 1.0, "naucs_at_100_std": 1.0, "naucs_at_100_diff1": 1.0, "naucs_at_1000_max": NaN, "naucs_at_1000_std": NaN, "naucs_at_1000_diff1": NaN}, "tatdqa": {"ndcg_at_1": 0.29645, "ndcg_at_3": 0.39956, "ndcg_at_5": 0.43827, "ndcg_at_10": 0.47478, "ndcg_at_20": 0.49868, "ndcg_at_100": 0.53079, "ndcg_at_1000": 0.53791, "map_at_1": 0.29645, "map_at_3": 0.37392, "map_at_5": 0.39545, "map_at_10": 0.41049, "map_at_20": 0.41714, "map_at_100": 0.42159, "map_at_1000": 0.42192, "recall_at_1": 0.29645, "recall_at_3": 0.47384, "recall_at_5": 0.56765, "recall_at_10": 0.6807, "recall_at_20": 0.7745, "recall_at_100": 0.94768, "recall_at_1000": 1.0, "precision_at_1": 0.29645, "precision_at_3": 0.15795, "precision_at_5": 0.11353, "precision_at_10": 0.06807, "precision_at_20": 0.03873, "precision_at_100": 0.00948, "precision_at_1000": 0.001, "mrr_at_1": 0.29224293445580274, "mrr_at_3": 0.3725195429945882, "mrr_at_5": 0.39368610944077, "mrr_at_10": 0.40873182907157685, "mrr_at_20": 0.4153525395345871, "mrr_at_100": 0.41982577879680105, "mrr_at_1000": 0.4201573230533317, "naucs_at_1_max": 0.26117941273429796, "naucs_at_1_std": -0.055826804856939846, "naucs_at_1_diff1": 0.5147725113819512, "naucs_at_3_max": 0.29085989220012026, "naucs_at_3_std": 0.009298164013586034, "naucs_at_3_diff1": 0.3875656176159266, "naucs_at_5_max": 0.33630142572206806, "naucs_at_5_std": 0.07004327823851751, "naucs_at_5_diff1": 0.3737553907577849, "naucs_at_10_max": 0.4217644421117308, "naucs_at_10_std": 0.1857452655534678, "naucs_at_10_diff1": 0.3593449086590725, "naucs_at_20_max": 0.44042964320456096, "naucs_at_20_std": 0.22640537350501502, "naucs_at_20_diff1": 0.3600417075223415, "naucs_at_100_max": 0.6593118291876395, "naucs_at_100_std": 0.5650726504305896, "naucs_at_100_diff1": 0.4648047295341125, "naucs_at_1000_max": NaN, "naucs_at_1000_std": NaN, "naucs_at_1000_diff1": NaN}, "shift_project": {"ndcg_at_1": 0.4, "ndcg_at_3": 0.52333, "ndcg_at_5": 0.56121, "ndcg_at_10": 0.58406, "ndcg_at_20": 0.59639, "ndcg_at_100": 0.62271, "ndcg_at_1000": 0.62801, "map_at_1": 0.4, "map_at_3": 0.49333, "map_at_5": 0.51483, "map_at_10": 0.52438, "map_at_20": 0.52761, "map_at_100": 0.53134, "map_at_1000": 0.53156, "recall_at_1": 0.4, "recall_at_3": 0.61, "recall_at_5": 0.7, "recall_at_10": 0.77, "recall_at_20": 0.82, "recall_at_100": 0.96, "recall_at_1000": 1.0, "precision_at_1": 0.4, "precision_at_3": 0.20333, "precision_at_5": 0.14, "precision_at_10": 0.077, "precision_at_20": 0.041, "precision_at_100": 0.0096, "precision_at_1000": 0.001, "mrr_at_1": 0.4, "mrr_at_3": 0.4966666666666668, "mrr_at_5": 0.5191666666666668, "mrr_at_10": 0.5260396825396827, "mrr_at_20": 0.5293352290983873, "mrr_at_100": 0.5331134772704517, "mrr_at_1000": 0.5333385894468862, "naucs_at_1_max": 0.2815010298661175, "naucs_at_1_std": 0.11077497425334709, "naucs_at_1_diff1": 0.4076081359423273, "naucs_at_3_max": 0.213693290831357, "naucs_at_3_std": -0.0029006088257063556, "naucs_at_3_diff1": 0.29765987340051087, "naucs_at_5_max": 0.3129526211493426, "naucs_at_5_std": -0.09102864348765899, "naucs_at_5_diff1": 0.15919654116375467, "naucs_at_10_max": 0.35921308833732496, "naucs_at_10_std": -0.20668740628007307, "naucs_at_10_diff1": 0.2815290615697954, "naucs_at_20_max": 0.3365734111686314, "naucs_at_20_std": -0.30763299922898824, "naucs_at_20_diff1": 0.367909461394427, "naucs_at_100_max": 0.3764005602240897, "naucs_at_100_std": -0.6464752567693711, "naucs_at_100_diff1": 0.576914098972922, "naucs_at_1000_max": NaN, "naucs_at_1000_std": NaN, "naucs_at_1000_diff1": NaN}}
special_tokens_map.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ {
4
+ "content": "<image>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ }
10
+ ],
11
+ "bos_token": {
12
+ "content": "<bos>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false
17
+ },
18
+ "eos_token": {
19
+ "content": "<eos>",
20
+ "lstrip": false,
21
+ "normalized": false,
22
+ "rstrip": false,
23
+ "single_word": false
24
+ },
25
+ "pad_token": {
26
+ "content": "<pad>",
27
+ "lstrip": false,
28
+ "normalized": false,
29
+ "rstrip": false,
30
+ "single_word": false
31
+ },
32
+ "unk_token": {
33
+ "content": "<unk>",
34
+ "lstrip": false,
35
+ "normalized": false,
36
+ "rstrip": false,
37
+ "single_word": false
38
+ }
39
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ffd310e50986db7a039948ab83441d612689e7f989198e31b5c8984ca458adf6
3
+ size 17763459
tokenizer_config.json ADDED
The diff for this file is too large to render. See raw diff
 
training_config.yml ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ config:
2
+ (): colpali_engine.utils.train_colpali_engine_models.ColModelTrainingConfig
3
+ output_dir: !path ../../../models/right_pad/train_bipali_pairwise_hardneg
4
+ processor:
5
+ () : colpali_engine.utils.wrapper.AutoProcessorWrapper
6
+ pretrained_model_name_or_path: "./models/paligemma-3b-mix-448"
7
+ max_length: 50
8
+ model:
9
+ (): colpali_engine.utils.wrapper.AllPurposeWrapper
10
+ class_to_instanciate: !ext colpali_engine.models.paligemma_colbert_architecture.BiPali
11
+ pretrained_model_name_or_path: "./models/paligemma-3b-mix-448"
12
+ torch_dtype: !ext torch.bfloat16
13
+ # device_map: "auto"
14
+ # quantization_config:
15
+ # (): transformers.BitsAndBytesConfig
16
+ # load_in_4bit: true
17
+ # bnb_4bit_quant_type: "nf4"
18
+ # bnb_4bit_compute_dtype: "bfloat16"
19
+ # bnb_4bit_use_double_quant: true
20
+
21
+ dataset_loading_func: !ext colpali_engine.utils.dataset_transformation.load_train_set_ir_negs
22
+ eval_dataset_loader: !import ../data/test_data.yaml
23
+
24
+ max_length: 50
25
+ run_eval: true
26
+ add_suffix: true
27
+ loss_func:
28
+ (): colpali_engine.loss.colbert_loss.BiPairwiseNegativeCELoss
29
+ in_batch_term: true
30
+ tr_args: !import ../tr_args/default_neg_tr_args.yaml
31
+ peft_config:
32
+ (): peft.LoraConfig
33
+ r: 32
34
+ lora_alpha: 32
35
+ lora_dropout: 0.1
36
+ init_lora_weights: "gaussian"
37
+ bias: "none"
38
+ task_type: "FEATURE_EXTRACTION"
39
+ target_modules: '(.*(language_model).*(down_proj|gate_proj|up_proj|k_proj|q_proj|v_proj|o_proj).*$)'
40
+ # target_modules: '(.*(language_model).*(down_proj|gate_proj|up_proj|k_proj|q_proj|v_proj|o_proj).*$'
41
+