ColPali
Safetensors
English
vidore
vidore-experimental
manu commited on
Commit
7102729
1 Parent(s): 19e5fe0

Upload folder using huggingface_hub

Browse files
README.md ADDED
@@ -0,0 +1,202 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: vidore/colqwen2-base
3
+ library_name: peft
4
+ ---
5
+
6
+ # Model Card for Model ID
7
+
8
+ <!-- Provide a quick summary of what the model is/does. -->
9
+
10
+
11
+
12
+ ## Model Details
13
+
14
+ ### Model Description
15
+
16
+ <!-- Provide a longer summary of what this model is. -->
17
+
18
+
19
+
20
+ - **Developed by:** [More Information Needed]
21
+ - **Funded by [optional]:** [More Information Needed]
22
+ - **Shared by [optional]:** [More Information Needed]
23
+ - **Model type:** [More Information Needed]
24
+ - **Language(s) (NLP):** [More Information Needed]
25
+ - **License:** [More Information Needed]
26
+ - **Finetuned from model [optional]:** [More Information Needed]
27
+
28
+ ### Model Sources [optional]
29
+
30
+ <!-- Provide the basic links for the model. -->
31
+
32
+ - **Repository:** [More Information Needed]
33
+ - **Paper [optional]:** [More Information Needed]
34
+ - **Demo [optional]:** [More Information Needed]
35
+
36
+ ## Uses
37
+
38
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
39
+
40
+ ### Direct Use
41
+
42
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
43
+
44
+ [More Information Needed]
45
+
46
+ ### Downstream Use [optional]
47
+
48
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
49
+
50
+ [More Information Needed]
51
+
52
+ ### Out-of-Scope Use
53
+
54
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
55
+
56
+ [More Information Needed]
57
+
58
+ ## Bias, Risks, and Limitations
59
+
60
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
61
+
62
+ [More Information Needed]
63
+
64
+ ### Recommendations
65
+
66
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
67
+
68
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
69
+
70
+ ## How to Get Started with the Model
71
+
72
+ Use the code below to get started with the model.
73
+
74
+ [More Information Needed]
75
+
76
+ ## Training Details
77
+
78
+ ### Training Data
79
+
80
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
81
+
82
+ [More Information Needed]
83
+
84
+ ### Training Procedure
85
+
86
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
87
+
88
+ #### Preprocessing [optional]
89
+
90
+ [More Information Needed]
91
+
92
+
93
+ #### Training Hyperparameters
94
+
95
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
96
+
97
+ #### Speeds, Sizes, Times [optional]
98
+
99
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
100
+
101
+ [More Information Needed]
102
+
103
+ ## Evaluation
104
+
105
+ <!-- This section describes the evaluation protocols and provides the results. -->
106
+
107
+ ### Testing Data, Factors & Metrics
108
+
109
+ #### Testing Data
110
+
111
+ <!-- This should link to a Dataset Card if possible. -->
112
+
113
+ [More Information Needed]
114
+
115
+ #### Factors
116
+
117
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
118
+
119
+ [More Information Needed]
120
+
121
+ #### Metrics
122
+
123
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
124
+
125
+ [More Information Needed]
126
+
127
+ ### Results
128
+
129
+ [More Information Needed]
130
+
131
+ #### Summary
132
+
133
+
134
+
135
+ ## Model Examination [optional]
136
+
137
+ <!-- Relevant interpretability work for the model goes here -->
138
+
139
+ [More Information Needed]
140
+
141
+ ## Environmental Impact
142
+
143
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
144
+
145
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
146
+
147
+ - **Hardware Type:** [More Information Needed]
148
+ - **Hours used:** [More Information Needed]
149
+ - **Cloud Provider:** [More Information Needed]
150
+ - **Compute Region:** [More Information Needed]
151
+ - **Carbon Emitted:** [More Information Needed]
152
+
153
+ ## Technical Specifications [optional]
154
+
155
+ ### Model Architecture and Objective
156
+
157
+ [More Information Needed]
158
+
159
+ ### Compute Infrastructure
160
+
161
+ [More Information Needed]
162
+
163
+ #### Hardware
164
+
165
+ [More Information Needed]
166
+
167
+ #### Software
168
+
169
+ [More Information Needed]
170
+
171
+ ## Citation [optional]
172
+
173
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
174
+
175
+ **BibTeX:**
176
+
177
+ [More Information Needed]
178
+
179
+ **APA:**
180
+
181
+ [More Information Needed]
182
+
183
+ ## Glossary [optional]
184
+
185
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
186
+
187
+ [More Information Needed]
188
+
189
+ ## More Information [optional]
190
+
191
+ [More Information Needed]
192
+
193
+ ## Model Card Authors [optional]
194
+
195
+ [More Information Needed]
196
+
197
+ ## Model Card Contact
198
+
199
+ [More Information Needed]
200
+ ### Framework versions
201
+
202
+ - PEFT 0.11.1
adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "vidore/colqwen2-base",
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": "(.*(model).*(down_proj|gate_proj|up_proj|k_proj|q_proj|v_proj|o_proj).*$|.*(custom_text_proj).*$)",
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 sha256:1f26000fd1fbd64be94a32e99012afde7b375f8342bb54b7a7dd0c1dd0d57066
3
+ size 74018232
added_tokens.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "<|box_end|>": 151649,
3
+ "<|box_start|>": 151648,
4
+ "<|endoftext|>": 151643,
5
+ "<|im_end|>": 151645,
6
+ "<|im_start|>": 151644,
7
+ "<|image_pad|>": 151655,
8
+ "<|object_ref_end|>": 151647,
9
+ "<|object_ref_start|>": 151646,
10
+ "<|quad_end|>": 151651,
11
+ "<|quad_start|>": 151650,
12
+ "<|video_pad|>": 151656,
13
+ "<|vision_end|>": 151653,
14
+ "<|vision_pad|>": 151654,
15
+ "<|vision_start|>": 151652
16
+ }
chat_template.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "chat_template": "{% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n{% endif %}<|im_start|>{{ message['role'] }}\n{% if message['content'] is string %}{{ message['content'] }}<|im_end|>\n{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>\n{% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant\n{% endif %}"
3
+ }
checkpoint-3694/README.md ADDED
@@ -0,0 +1,202 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: ./models/colqwen2_base
3
+ library_name: peft
4
+ ---
5
+
6
+ # Model Card for Model ID
7
+
8
+ <!-- Provide a quick summary of what the model is/does. -->
9
+
10
+
11
+
12
+ ## Model Details
13
+
14
+ ### Model Description
15
+
16
+ <!-- Provide a longer summary of what this model is. -->
17
+
18
+
19
+
20
+ - **Developed by:** [More Information Needed]
21
+ - **Funded by [optional]:** [More Information Needed]
22
+ - **Shared by [optional]:** [More Information Needed]
23
+ - **Model type:** [More Information Needed]
24
+ - **Language(s) (NLP):** [More Information Needed]
25
+ - **License:** [More Information Needed]
26
+ - **Finetuned from model [optional]:** [More Information Needed]
27
+
28
+ ### Model Sources [optional]
29
+
30
+ <!-- Provide the basic links for the model. -->
31
+
32
+ - **Repository:** [More Information Needed]
33
+ - **Paper [optional]:** [More Information Needed]
34
+ - **Demo [optional]:** [More Information Needed]
35
+
36
+ ## Uses
37
+
38
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
39
+
40
+ ### Direct Use
41
+
42
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
43
+
44
+ [More Information Needed]
45
+
46
+ ### Downstream Use [optional]
47
+
48
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
49
+
50
+ [More Information Needed]
51
+
52
+ ### Out-of-Scope Use
53
+
54
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
55
+
56
+ [More Information Needed]
57
+
58
+ ## Bias, Risks, and Limitations
59
+
60
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
61
+
62
+ [More Information Needed]
63
+
64
+ ### Recommendations
65
+
66
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
67
+
68
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
69
+
70
+ ## How to Get Started with the Model
71
+
72
+ Use the code below to get started with the model.
73
+
74
+ [More Information Needed]
75
+
76
+ ## Training Details
77
+
78
+ ### Training Data
79
+
80
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
81
+
82
+ [More Information Needed]
83
+
84
+ ### Training Procedure
85
+
86
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
87
+
88
+ #### Preprocessing [optional]
89
+
90
+ [More Information Needed]
91
+
92
+
93
+ #### Training Hyperparameters
94
+
95
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
96
+
97
+ #### Speeds, Sizes, Times [optional]
98
+
99
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
100
+
101
+ [More Information Needed]
102
+
103
+ ## Evaluation
104
+
105
+ <!-- This section describes the evaluation protocols and provides the results. -->
106
+
107
+ ### Testing Data, Factors & Metrics
108
+
109
+ #### Testing Data
110
+
111
+ <!-- This should link to a Dataset Card if possible. -->
112
+
113
+ [More Information Needed]
114
+
115
+ #### Factors
116
+
117
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
118
+
119
+ [More Information Needed]
120
+
121
+ #### Metrics
122
+
123
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
124
+
125
+ [More Information Needed]
126
+
127
+ ### Results
128
+
129
+ [More Information Needed]
130
+
131
+ #### Summary
132
+
133
+
134
+
135
+ ## Model Examination [optional]
136
+
137
+ <!-- Relevant interpretability work for the model goes here -->
138
+
139
+ [More Information Needed]
140
+
141
+ ## Environmental Impact
142
+
143
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
144
+
145
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
146
+
147
+ - **Hardware Type:** [More Information Needed]
148
+ - **Hours used:** [More Information Needed]
149
+ - **Cloud Provider:** [More Information Needed]
150
+ - **Compute Region:** [More Information Needed]
151
+ - **Carbon Emitted:** [More Information Needed]
152
+
153
+ ## Technical Specifications [optional]
154
+
155
+ ### Model Architecture and Objective
156
+
157
+ [More Information Needed]
158
+
159
+ ### Compute Infrastructure
160
+
161
+ [More Information Needed]
162
+
163
+ #### Hardware
164
+
165
+ [More Information Needed]
166
+
167
+ #### Software
168
+
169
+ [More Information Needed]
170
+
171
+ ## Citation [optional]
172
+
173
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
174
+
175
+ **BibTeX:**
176
+
177
+ [More Information Needed]
178
+
179
+ **APA:**
180
+
181
+ [More Information Needed]
182
+
183
+ ## Glossary [optional]
184
+
185
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
186
+
187
+ [More Information Needed]
188
+
189
+ ## More Information [optional]
190
+
191
+ [More Information Needed]
192
+
193
+ ## Model Card Authors [optional]
194
+
195
+ [More Information Needed]
196
+
197
+ ## Model Card Contact
198
+
199
+ [More Information Needed]
200
+ ### Framework versions
201
+
202
+ - PEFT 0.11.1
checkpoint-3694/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "./models/colqwen2_base",
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": "(.*(model).*(down_proj|gate_proj|up_proj|k_proj|q_proj|v_proj|o_proj).*$|.*(custom_text_proj).*$)",
23
+ "task_type": "FEATURE_EXTRACTION",
24
+ "use_dora": false,
25
+ "use_rslora": false
26
+ }
checkpoint-3694/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1f26000fd1fbd64be94a32e99012afde7b375f8342bb54b7a7dd0c1dd0d57066
3
+ size 74018232
checkpoint-3694/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:94dfe0075ec4f21387ac18e77c648bfc5db314e088bdcdb8d4fdf6e30532496b
3
+ size 148262384
checkpoint-3694/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:75e9665951c5d547a5de8ed05c5981901bf7eeb728786a0acc4cf999aa58e092
3
+ size 14244
checkpoint-3694/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f0c4c047aa3abb1bc366c2fb0e6a4fd601029da0338c37929023a3ff5f04bef3
3
+ size 1064
checkpoint-3694/trainer_state.json ADDED
@@ -0,0 +1,2904 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "eval_steps": 100,
6
+ "global_step": 3694,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0027070925825663237,
13
+ "grad_norm": 5.65625,
14
+ "learning_rate": 1.0000000000000002e-06,
15
+ "loss": 0.7343,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.005414185165132647,
20
+ "grad_norm": 8.9375,
21
+ "learning_rate": 2.0000000000000003e-06,
22
+ "loss": 0.864,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.008121277747698972,
27
+ "grad_norm": 9.0625,
28
+ "learning_rate": 3e-06,
29
+ "loss": 0.7618,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.010828370330265295,
34
+ "grad_norm": 7.375,
35
+ "learning_rate": 4.000000000000001e-06,
36
+ "loss": 0.749,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.01353546291283162,
41
+ "grad_norm": 5.4375,
42
+ "learning_rate": 5e-06,
43
+ "loss": 0.7434,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.016242555495397944,
48
+ "grad_norm": 5.34375,
49
+ "learning_rate": 6e-06,
50
+ "loss": 0.7079,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.018949648077964266,
55
+ "grad_norm": 4.25,
56
+ "learning_rate": 7.000000000000001e-06,
57
+ "loss": 0.7368,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.02165674066053059,
62
+ "grad_norm": 3.828125,
63
+ "learning_rate": 8.000000000000001e-06,
64
+ "loss": 0.682,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.024363833243096916,
69
+ "grad_norm": 2.875,
70
+ "learning_rate": 9e-06,
71
+ "loss": 0.6489,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.02707092582566324,
76
+ "grad_norm": 2.453125,
77
+ "learning_rate": 1e-05,
78
+ "loss": 0.589,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.02707092582566324,
83
+ "eval_loss": 0.6064066290855408,
84
+ "eval_runtime": 155.4734,
85
+ "eval_samples_per_second": 3.293,
86
+ "eval_steps_per_second": 0.103,
87
+ "step": 100
88
+ },
89
+ {
90
+ "epoch": 0.02977801840822956,
91
+ "grad_norm": 2.1875,
92
+ "learning_rate": 1.1000000000000001e-05,
93
+ "loss": 0.5723,
94
+ "step": 110
95
+ },
96
+ {
97
+ "epoch": 0.03248511099079589,
98
+ "grad_norm": 3.640625,
99
+ "learning_rate": 1.2e-05,
100
+ "loss": 0.5525,
101
+ "step": 120
102
+ },
103
+ {
104
+ "epoch": 0.03519220357336221,
105
+ "grad_norm": 2.34375,
106
+ "learning_rate": 1.3000000000000001e-05,
107
+ "loss": 0.5503,
108
+ "step": 130
109
+ },
110
+ {
111
+ "epoch": 0.03789929615592853,
112
+ "grad_norm": 2.3125,
113
+ "learning_rate": 1.4000000000000001e-05,
114
+ "loss": 0.4926,
115
+ "step": 140
116
+ },
117
+ {
118
+ "epoch": 0.040606388738494856,
119
+ "grad_norm": 2.578125,
120
+ "learning_rate": 1.5e-05,
121
+ "loss": 0.4831,
122
+ "step": 150
123
+ },
124
+ {
125
+ "epoch": 0.04331348132106118,
126
+ "grad_norm": 2.109375,
127
+ "learning_rate": 1.6000000000000003e-05,
128
+ "loss": 0.4413,
129
+ "step": 160
130
+ },
131
+ {
132
+ "epoch": 0.0460205739036275,
133
+ "grad_norm": 2.671875,
134
+ "learning_rate": 1.7000000000000003e-05,
135
+ "loss": 0.4562,
136
+ "step": 170
137
+ },
138
+ {
139
+ "epoch": 0.04872766648619383,
140
+ "grad_norm": 2.203125,
141
+ "learning_rate": 1.8e-05,
142
+ "loss": 0.4348,
143
+ "step": 180
144
+ },
145
+ {
146
+ "epoch": 0.051434759068760154,
147
+ "grad_norm": 2.015625,
148
+ "learning_rate": 1.9e-05,
149
+ "loss": 0.4386,
150
+ "step": 190
151
+ },
152
+ {
153
+ "epoch": 0.05414185165132648,
154
+ "grad_norm": 2.09375,
155
+ "learning_rate": 2e-05,
156
+ "loss": 0.4009,
157
+ "step": 200
158
+ },
159
+ {
160
+ "epoch": 0.05414185165132648,
161
+ "eval_loss": 0.4071368873119354,
162
+ "eval_runtime": 104.5667,
163
+ "eval_samples_per_second": 4.896,
164
+ "eval_steps_per_second": 0.153,
165
+ "step": 200
166
+ },
167
+ {
168
+ "epoch": 0.0568489442338928,
169
+ "grad_norm": 2.25,
170
+ "learning_rate": 2.1e-05,
171
+ "loss": 0.4026,
172
+ "step": 210
173
+ },
174
+ {
175
+ "epoch": 0.05955603681645912,
176
+ "grad_norm": 1.90625,
177
+ "learning_rate": 2.2000000000000003e-05,
178
+ "loss": 0.3774,
179
+ "step": 220
180
+ },
181
+ {
182
+ "epoch": 0.062263129399025445,
183
+ "grad_norm": 1.8828125,
184
+ "learning_rate": 2.3000000000000003e-05,
185
+ "loss": 0.3734,
186
+ "step": 230
187
+ },
188
+ {
189
+ "epoch": 0.06497022198159177,
190
+ "grad_norm": 1.5703125,
191
+ "learning_rate": 2.4e-05,
192
+ "loss": 0.331,
193
+ "step": 240
194
+ },
195
+ {
196
+ "epoch": 0.0676773145641581,
197
+ "grad_norm": 1.671875,
198
+ "learning_rate": 2.5e-05,
199
+ "loss": 0.3381,
200
+ "step": 250
201
+ },
202
+ {
203
+ "epoch": 0.07038440714672442,
204
+ "grad_norm": 1.8515625,
205
+ "learning_rate": 2.6000000000000002e-05,
206
+ "loss": 0.3392,
207
+ "step": 260
208
+ },
209
+ {
210
+ "epoch": 0.07309149972929074,
211
+ "grad_norm": 1.625,
212
+ "learning_rate": 2.7000000000000002e-05,
213
+ "loss": 0.3165,
214
+ "step": 270
215
+ },
216
+ {
217
+ "epoch": 0.07579859231185707,
218
+ "grad_norm": 1.671875,
219
+ "learning_rate": 2.8000000000000003e-05,
220
+ "loss": 0.3162,
221
+ "step": 280
222
+ },
223
+ {
224
+ "epoch": 0.07850568489442339,
225
+ "grad_norm": 1.953125,
226
+ "learning_rate": 2.9e-05,
227
+ "loss": 0.2844,
228
+ "step": 290
229
+ },
230
+ {
231
+ "epoch": 0.08121277747698971,
232
+ "grad_norm": 2.28125,
233
+ "learning_rate": 3e-05,
234
+ "loss": 0.3093,
235
+ "step": 300
236
+ },
237
+ {
238
+ "epoch": 0.08121277747698971,
239
+ "eval_loss": 0.2839646637439728,
240
+ "eval_runtime": 103.4698,
241
+ "eval_samples_per_second": 4.948,
242
+ "eval_steps_per_second": 0.155,
243
+ "step": 300
244
+ },
245
+ {
246
+ "epoch": 0.08391987005955603,
247
+ "grad_norm": 1.4296875,
248
+ "learning_rate": 3.1e-05,
249
+ "loss": 0.2753,
250
+ "step": 310
251
+ },
252
+ {
253
+ "epoch": 0.08662696264212236,
254
+ "grad_norm": 1.4609375,
255
+ "learning_rate": 3.2000000000000005e-05,
256
+ "loss": 0.2625,
257
+ "step": 320
258
+ },
259
+ {
260
+ "epoch": 0.08933405522468868,
261
+ "grad_norm": 1.5390625,
262
+ "learning_rate": 3.3e-05,
263
+ "loss": 0.2525,
264
+ "step": 330
265
+ },
266
+ {
267
+ "epoch": 0.092041147807255,
268
+ "grad_norm": 1.9453125,
269
+ "learning_rate": 3.4000000000000007e-05,
270
+ "loss": 0.2897,
271
+ "step": 340
272
+ },
273
+ {
274
+ "epoch": 0.09474824038982133,
275
+ "grad_norm": 1.4453125,
276
+ "learning_rate": 3.5e-05,
277
+ "loss": 0.243,
278
+ "step": 350
279
+ },
280
+ {
281
+ "epoch": 0.09745533297238766,
282
+ "grad_norm": 1.6484375,
283
+ "learning_rate": 3.6e-05,
284
+ "loss": 0.2442,
285
+ "step": 360
286
+ },
287
+ {
288
+ "epoch": 0.10016242555495398,
289
+ "grad_norm": 1.15625,
290
+ "learning_rate": 3.7e-05,
291
+ "loss": 0.2139,
292
+ "step": 370
293
+ },
294
+ {
295
+ "epoch": 0.10286951813752031,
296
+ "grad_norm": 1.3203125,
297
+ "learning_rate": 3.8e-05,
298
+ "loss": 0.2312,
299
+ "step": 380
300
+ },
301
+ {
302
+ "epoch": 0.10557661072008663,
303
+ "grad_norm": 2.703125,
304
+ "learning_rate": 3.9000000000000006e-05,
305
+ "loss": 0.2202,
306
+ "step": 390
307
+ },
308
+ {
309
+ "epoch": 0.10828370330265295,
310
+ "grad_norm": 1.078125,
311
+ "learning_rate": 4e-05,
312
+ "loss": 0.2115,
313
+ "step": 400
314
+ },
315
+ {
316
+ "epoch": 0.10828370330265295,
317
+ "eval_loss": 0.23121848702430725,
318
+ "eval_runtime": 134.1955,
319
+ "eval_samples_per_second": 3.815,
320
+ "eval_steps_per_second": 0.119,
321
+ "step": 400
322
+ },
323
+ {
324
+ "epoch": 0.11099079588521928,
325
+ "grad_norm": 1.828125,
326
+ "learning_rate": 4.1e-05,
327
+ "loss": 0.1976,
328
+ "step": 410
329
+ },
330
+ {
331
+ "epoch": 0.1136978884677856,
332
+ "grad_norm": 1.0546875,
333
+ "learning_rate": 4.2e-05,
334
+ "loss": 0.2182,
335
+ "step": 420
336
+ },
337
+ {
338
+ "epoch": 0.11640498105035192,
339
+ "grad_norm": 1.1015625,
340
+ "learning_rate": 4.3e-05,
341
+ "loss": 0.1767,
342
+ "step": 430
343
+ },
344
+ {
345
+ "epoch": 0.11911207363291824,
346
+ "grad_norm": 3.078125,
347
+ "learning_rate": 4.4000000000000006e-05,
348
+ "loss": 0.2058,
349
+ "step": 440
350
+ },
351
+ {
352
+ "epoch": 0.12181916621548457,
353
+ "grad_norm": 2.25,
354
+ "learning_rate": 4.5e-05,
355
+ "loss": 0.168,
356
+ "step": 450
357
+ },
358
+ {
359
+ "epoch": 0.12452625879805089,
360
+ "grad_norm": 2.5625,
361
+ "learning_rate": 4.600000000000001e-05,
362
+ "loss": 0.1728,
363
+ "step": 460
364
+ },
365
+ {
366
+ "epoch": 0.12723335138061723,
367
+ "grad_norm": 2.109375,
368
+ "learning_rate": 4.7e-05,
369
+ "loss": 0.1927,
370
+ "step": 470
371
+ },
372
+ {
373
+ "epoch": 0.12994044396318355,
374
+ "grad_norm": 2.140625,
375
+ "learning_rate": 4.8e-05,
376
+ "loss": 0.1433,
377
+ "step": 480
378
+ },
379
+ {
380
+ "epoch": 0.13264753654574987,
381
+ "grad_norm": 1.6640625,
382
+ "learning_rate": 4.9e-05,
383
+ "loss": 0.1728,
384
+ "step": 490
385
+ },
386
+ {
387
+ "epoch": 0.1353546291283162,
388
+ "grad_norm": 2.4375,
389
+ "learning_rate": 5e-05,
390
+ "loss": 0.176,
391
+ "step": 500
392
+ },
393
+ {
394
+ "epoch": 0.1353546291283162,
395
+ "eval_loss": 0.19780531525611877,
396
+ "eval_runtime": 102.6349,
397
+ "eval_samples_per_second": 4.989,
398
+ "eval_steps_per_second": 0.156,
399
+ "step": 500
400
+ },
401
+ {
402
+ "epoch": 0.13806172171088252,
403
+ "grad_norm": 1.8984375,
404
+ "learning_rate": 4.984345648090169e-05,
405
+ "loss": 0.1639,
406
+ "step": 510
407
+ },
408
+ {
409
+ "epoch": 0.14076881429344884,
410
+ "grad_norm": 1.5078125,
411
+ "learning_rate": 4.9686912961803384e-05,
412
+ "loss": 0.1598,
413
+ "step": 520
414
+ },
415
+ {
416
+ "epoch": 0.14347590687601516,
417
+ "grad_norm": 5.625,
418
+ "learning_rate": 4.9530369442705075e-05,
419
+ "loss": 0.17,
420
+ "step": 530
421
+ },
422
+ {
423
+ "epoch": 0.1461829994585815,
424
+ "grad_norm": 2.390625,
425
+ "learning_rate": 4.9373825923606765e-05,
426
+ "loss": 0.1601,
427
+ "step": 540
428
+ },
429
+ {
430
+ "epoch": 0.1488900920411478,
431
+ "grad_norm": 2.28125,
432
+ "learning_rate": 4.9217282404508456e-05,
433
+ "loss": 0.2067,
434
+ "step": 550
435
+ },
436
+ {
437
+ "epoch": 0.15159718462371413,
438
+ "grad_norm": 1.296875,
439
+ "learning_rate": 4.906073888541015e-05,
440
+ "loss": 0.1396,
441
+ "step": 560
442
+ },
443
+ {
444
+ "epoch": 0.15430427720628045,
445
+ "grad_norm": 2.3125,
446
+ "learning_rate": 4.890419536631184e-05,
447
+ "loss": 0.1379,
448
+ "step": 570
449
+ },
450
+ {
451
+ "epoch": 0.15701136978884678,
452
+ "grad_norm": 1.984375,
453
+ "learning_rate": 4.874765184721353e-05,
454
+ "loss": 0.1722,
455
+ "step": 580
456
+ },
457
+ {
458
+ "epoch": 0.1597184623714131,
459
+ "grad_norm": 2.796875,
460
+ "learning_rate": 4.859110832811522e-05,
461
+ "loss": 0.1608,
462
+ "step": 590
463
+ },
464
+ {
465
+ "epoch": 0.16242555495397942,
466
+ "grad_norm": 2.4375,
467
+ "learning_rate": 4.843456480901691e-05,
468
+ "loss": 0.1576,
469
+ "step": 600
470
+ },
471
+ {
472
+ "epoch": 0.16242555495397942,
473
+ "eval_loss": 0.1739654242992401,
474
+ "eval_runtime": 102.4755,
475
+ "eval_samples_per_second": 4.996,
476
+ "eval_steps_per_second": 0.156,
477
+ "step": 600
478
+ },
479
+ {
480
+ "epoch": 0.16513264753654575,
481
+ "grad_norm": 1.78125,
482
+ "learning_rate": 4.82780212899186e-05,
483
+ "loss": 0.1292,
484
+ "step": 610
485
+ },
486
+ {
487
+ "epoch": 0.16783974011911207,
488
+ "grad_norm": 2.296875,
489
+ "learning_rate": 4.812147777082029e-05,
490
+ "loss": 0.1421,
491
+ "step": 620
492
+ },
493
+ {
494
+ "epoch": 0.1705468327016784,
495
+ "grad_norm": 3.53125,
496
+ "learning_rate": 4.796493425172198e-05,
497
+ "loss": 0.1473,
498
+ "step": 630
499
+ },
500
+ {
501
+ "epoch": 0.17325392528424471,
502
+ "grad_norm": 1.6015625,
503
+ "learning_rate": 4.780839073262367e-05,
504
+ "loss": 0.1313,
505
+ "step": 640
506
+ },
507
+ {
508
+ "epoch": 0.17596101786681104,
509
+ "grad_norm": 2.75,
510
+ "learning_rate": 4.765184721352536e-05,
511
+ "loss": 0.0987,
512
+ "step": 650
513
+ },
514
+ {
515
+ "epoch": 0.17866811044937736,
516
+ "grad_norm": 2.453125,
517
+ "learning_rate": 4.7495303694427054e-05,
518
+ "loss": 0.1301,
519
+ "step": 660
520
+ },
521
+ {
522
+ "epoch": 0.18137520303194368,
523
+ "grad_norm": 1.203125,
524
+ "learning_rate": 4.7338760175328744e-05,
525
+ "loss": 0.118,
526
+ "step": 670
527
+ },
528
+ {
529
+ "epoch": 0.18408229561451,
530
+ "grad_norm": 2.703125,
531
+ "learning_rate": 4.7182216656230435e-05,
532
+ "loss": 0.1384,
533
+ "step": 680
534
+ },
535
+ {
536
+ "epoch": 0.18678938819707633,
537
+ "grad_norm": 0.78515625,
538
+ "learning_rate": 4.7025673137132126e-05,
539
+ "loss": 0.0855,
540
+ "step": 690
541
+ },
542
+ {
543
+ "epoch": 0.18949648077964265,
544
+ "grad_norm": 2.34375,
545
+ "learning_rate": 4.6869129618033816e-05,
546
+ "loss": 0.148,
547
+ "step": 700
548
+ },
549
+ {
550
+ "epoch": 0.18949648077964265,
551
+ "eval_loss": 0.1687919795513153,
552
+ "eval_runtime": 102.502,
553
+ "eval_samples_per_second": 4.995,
554
+ "eval_steps_per_second": 0.156,
555
+ "step": 700
556
+ },
557
+ {
558
+ "epoch": 0.19220357336220897,
559
+ "grad_norm": 0.81640625,
560
+ "learning_rate": 4.671258609893551e-05,
561
+ "loss": 0.134,
562
+ "step": 710
563
+ },
564
+ {
565
+ "epoch": 0.19491066594477532,
566
+ "grad_norm": 1.3046875,
567
+ "learning_rate": 4.65560425798372e-05,
568
+ "loss": 0.1222,
569
+ "step": 720
570
+ },
571
+ {
572
+ "epoch": 0.19761775852734165,
573
+ "grad_norm": 2.5625,
574
+ "learning_rate": 4.639949906073889e-05,
575
+ "loss": 0.1511,
576
+ "step": 730
577
+ },
578
+ {
579
+ "epoch": 0.20032485110990797,
580
+ "grad_norm": 0.55078125,
581
+ "learning_rate": 4.624295554164057e-05,
582
+ "loss": 0.1661,
583
+ "step": 740
584
+ },
585
+ {
586
+ "epoch": 0.2030319436924743,
587
+ "grad_norm": 1.9140625,
588
+ "learning_rate": 4.608641202254227e-05,
589
+ "loss": 0.1381,
590
+ "step": 750
591
+ },
592
+ {
593
+ "epoch": 0.20573903627504062,
594
+ "grad_norm": 2.015625,
595
+ "learning_rate": 4.5929868503443954e-05,
596
+ "loss": 0.1499,
597
+ "step": 760
598
+ },
599
+ {
600
+ "epoch": 0.20844612885760694,
601
+ "grad_norm": 2.265625,
602
+ "learning_rate": 4.577332498434565e-05,
603
+ "loss": 0.1692,
604
+ "step": 770
605
+ },
606
+ {
607
+ "epoch": 0.21115322144017326,
608
+ "grad_norm": 2.40625,
609
+ "learning_rate": 4.561678146524734e-05,
610
+ "loss": 0.1226,
611
+ "step": 780
612
+ },
613
+ {
614
+ "epoch": 0.21386031402273958,
615
+ "grad_norm": 2.484375,
616
+ "learning_rate": 4.546023794614903e-05,
617
+ "loss": 0.1594,
618
+ "step": 790
619
+ },
620
+ {
621
+ "epoch": 0.2165674066053059,
622
+ "grad_norm": 0.76171875,
623
+ "learning_rate": 4.5303694427050724e-05,
624
+ "loss": 0.0962,
625
+ "step": 800
626
+ },
627
+ {
628
+ "epoch": 0.2165674066053059,
629
+ "eval_loss": 0.15033257007598877,
630
+ "eval_runtime": 102.8567,
631
+ "eval_samples_per_second": 4.978,
632
+ "eval_steps_per_second": 0.156,
633
+ "step": 800
634
+ },
635
+ {
636
+ "epoch": 0.21927449918787223,
637
+ "grad_norm": 1.4453125,
638
+ "learning_rate": 4.5147150907952414e-05,
639
+ "loss": 0.1581,
640
+ "step": 810
641
+ },
642
+ {
643
+ "epoch": 0.22198159177043855,
644
+ "grad_norm": 2.078125,
645
+ "learning_rate": 4.4990607388854105e-05,
646
+ "loss": 0.155,
647
+ "step": 820
648
+ },
649
+ {
650
+ "epoch": 0.22468868435300487,
651
+ "grad_norm": 0.921875,
652
+ "learning_rate": 4.4834063869755796e-05,
653
+ "loss": 0.1103,
654
+ "step": 830
655
+ },
656
+ {
657
+ "epoch": 0.2273957769355712,
658
+ "grad_norm": 0.90234375,
659
+ "learning_rate": 4.4677520350657486e-05,
660
+ "loss": 0.1323,
661
+ "step": 840
662
+ },
663
+ {
664
+ "epoch": 0.23010286951813752,
665
+ "grad_norm": 0.953125,
666
+ "learning_rate": 4.452097683155918e-05,
667
+ "loss": 0.1073,
668
+ "step": 850
669
+ },
670
+ {
671
+ "epoch": 0.23280996210070384,
672
+ "grad_norm": 0.69921875,
673
+ "learning_rate": 4.436443331246087e-05,
674
+ "loss": 0.1333,
675
+ "step": 860
676
+ },
677
+ {
678
+ "epoch": 0.23551705468327017,
679
+ "grad_norm": 4.34375,
680
+ "learning_rate": 4.420788979336256e-05,
681
+ "loss": 0.1429,
682
+ "step": 870
683
+ },
684
+ {
685
+ "epoch": 0.2382241472658365,
686
+ "grad_norm": 0.5859375,
687
+ "learning_rate": 4.405134627426425e-05,
688
+ "loss": 0.0988,
689
+ "step": 880
690
+ },
691
+ {
692
+ "epoch": 0.2409312398484028,
693
+ "grad_norm": 1.3046875,
694
+ "learning_rate": 4.389480275516594e-05,
695
+ "loss": 0.1184,
696
+ "step": 890
697
+ },
698
+ {
699
+ "epoch": 0.24363833243096913,
700
+ "grad_norm": 1.0234375,
701
+ "learning_rate": 4.373825923606763e-05,
702
+ "loss": 0.118,
703
+ "step": 900
704
+ },
705
+ {
706
+ "epoch": 0.24363833243096913,
707
+ "eval_loss": 0.14425741136074066,
708
+ "eval_runtime": 115.407,
709
+ "eval_samples_per_second": 4.436,
710
+ "eval_steps_per_second": 0.139,
711
+ "step": 900
712
+ },
713
+ {
714
+ "epoch": 0.24634542501353546,
715
+ "grad_norm": 1.34375,
716
+ "learning_rate": 4.358171571696932e-05,
717
+ "loss": 0.1032,
718
+ "step": 910
719
+ },
720
+ {
721
+ "epoch": 0.24905251759610178,
722
+ "grad_norm": 1.78125,
723
+ "learning_rate": 4.342517219787101e-05,
724
+ "loss": 0.1242,
725
+ "step": 920
726
+ },
727
+ {
728
+ "epoch": 0.2517596101786681,
729
+ "grad_norm": 1.296875,
730
+ "learning_rate": 4.3268628678772696e-05,
731
+ "loss": 0.131,
732
+ "step": 930
733
+ },
734
+ {
735
+ "epoch": 0.25446670276123445,
736
+ "grad_norm": 1.90625,
737
+ "learning_rate": 4.3112085159674393e-05,
738
+ "loss": 0.1058,
739
+ "step": 940
740
+ },
741
+ {
742
+ "epoch": 0.25717379534380075,
743
+ "grad_norm": 1.1015625,
744
+ "learning_rate": 4.295554164057608e-05,
745
+ "loss": 0.155,
746
+ "step": 950
747
+ },
748
+ {
749
+ "epoch": 0.2598808879263671,
750
+ "grad_norm": 1.2734375,
751
+ "learning_rate": 4.2798998121477775e-05,
752
+ "loss": 0.1103,
753
+ "step": 960
754
+ },
755
+ {
756
+ "epoch": 0.2625879805089334,
757
+ "grad_norm": 4.59375,
758
+ "learning_rate": 4.264245460237946e-05,
759
+ "loss": 0.1288,
760
+ "step": 970
761
+ },
762
+ {
763
+ "epoch": 0.26529507309149974,
764
+ "grad_norm": 0.90234375,
765
+ "learning_rate": 4.2485911083281156e-05,
766
+ "loss": 0.0967,
767
+ "step": 980
768
+ },
769
+ {
770
+ "epoch": 0.26800216567406604,
771
+ "grad_norm": 1.734375,
772
+ "learning_rate": 4.232936756418284e-05,
773
+ "loss": 0.1523,
774
+ "step": 990
775
+ },
776
+ {
777
+ "epoch": 0.2707092582566324,
778
+ "grad_norm": 2.90625,
779
+ "learning_rate": 4.217282404508454e-05,
780
+ "loss": 0.1364,
781
+ "step": 1000
782
+ },
783
+ {
784
+ "epoch": 0.2707092582566324,
785
+ "eval_loss": 0.14593610167503357,
786
+ "eval_runtime": 104.6862,
787
+ "eval_samples_per_second": 4.891,
788
+ "eval_steps_per_second": 0.153,
789
+ "step": 1000
790
+ },
791
+ {
792
+ "epoch": 0.2734163508391987,
793
+ "grad_norm": 2.125,
794
+ "learning_rate": 4.201628052598622e-05,
795
+ "loss": 0.1228,
796
+ "step": 1010
797
+ },
798
+ {
799
+ "epoch": 0.27612344342176504,
800
+ "grad_norm": 1.7265625,
801
+ "learning_rate": 4.185973700688792e-05,
802
+ "loss": 0.1114,
803
+ "step": 1020
804
+ },
805
+ {
806
+ "epoch": 0.27883053600433133,
807
+ "grad_norm": 1.0859375,
808
+ "learning_rate": 4.170319348778961e-05,
809
+ "loss": 0.1016,
810
+ "step": 1030
811
+ },
812
+ {
813
+ "epoch": 0.2815376285868977,
814
+ "grad_norm": 0.69921875,
815
+ "learning_rate": 4.15466499686913e-05,
816
+ "loss": 0.1023,
817
+ "step": 1040
818
+ },
819
+ {
820
+ "epoch": 0.284244721169464,
821
+ "grad_norm": 1.5859375,
822
+ "learning_rate": 4.139010644959299e-05,
823
+ "loss": 0.1097,
824
+ "step": 1050
825
+ },
826
+ {
827
+ "epoch": 0.2869518137520303,
828
+ "grad_norm": 0.5859375,
829
+ "learning_rate": 4.123356293049468e-05,
830
+ "loss": 0.1522,
831
+ "step": 1060
832
+ },
833
+ {
834
+ "epoch": 0.2896589063345966,
835
+ "grad_norm": 1.296875,
836
+ "learning_rate": 4.107701941139637e-05,
837
+ "loss": 0.1218,
838
+ "step": 1070
839
+ },
840
+ {
841
+ "epoch": 0.292365998917163,
842
+ "grad_norm": 1.9765625,
843
+ "learning_rate": 4.092047589229806e-05,
844
+ "loss": 0.1491,
845
+ "step": 1080
846
+ },
847
+ {
848
+ "epoch": 0.29507309149972927,
849
+ "grad_norm": 0.82421875,
850
+ "learning_rate": 4.0763932373199754e-05,
851
+ "loss": 0.1103,
852
+ "step": 1090
853
+ },
854
+ {
855
+ "epoch": 0.2977801840822956,
856
+ "grad_norm": 1.96875,
857
+ "learning_rate": 4.0607388854101445e-05,
858
+ "loss": 0.1419,
859
+ "step": 1100
860
+ },
861
+ {
862
+ "epoch": 0.2977801840822956,
863
+ "eval_loss": 0.14324142038822174,
864
+ "eval_runtime": 105.9997,
865
+ "eval_samples_per_second": 4.83,
866
+ "eval_steps_per_second": 0.151,
867
+ "step": 1100
868
+ },
869
+ {
870
+ "epoch": 0.3004872766648619,
871
+ "grad_norm": 0.8984375,
872
+ "learning_rate": 4.0450845335003135e-05,
873
+ "loss": 0.0784,
874
+ "step": 1110
875
+ },
876
+ {
877
+ "epoch": 0.30319436924742826,
878
+ "grad_norm": 0.9296875,
879
+ "learning_rate": 4.029430181590482e-05,
880
+ "loss": 0.0869,
881
+ "step": 1120
882
+ },
883
+ {
884
+ "epoch": 0.30590146182999456,
885
+ "grad_norm": 0.6015625,
886
+ "learning_rate": 4.013775829680652e-05,
887
+ "loss": 0.1135,
888
+ "step": 1130
889
+ },
890
+ {
891
+ "epoch": 0.3086085544125609,
892
+ "grad_norm": 1.4453125,
893
+ "learning_rate": 3.99812147777082e-05,
894
+ "loss": 0.1296,
895
+ "step": 1140
896
+ },
897
+ {
898
+ "epoch": 0.31131564699512726,
899
+ "grad_norm": 0.8203125,
900
+ "learning_rate": 3.98246712586099e-05,
901
+ "loss": 0.122,
902
+ "step": 1150
903
+ },
904
+ {
905
+ "epoch": 0.31402273957769355,
906
+ "grad_norm": 1.3203125,
907
+ "learning_rate": 3.966812773951158e-05,
908
+ "loss": 0.1102,
909
+ "step": 1160
910
+ },
911
+ {
912
+ "epoch": 0.3167298321602599,
913
+ "grad_norm": 0.98046875,
914
+ "learning_rate": 3.951158422041328e-05,
915
+ "loss": 0.1284,
916
+ "step": 1170
917
+ },
918
+ {
919
+ "epoch": 0.3194369247428262,
920
+ "grad_norm": 2.390625,
921
+ "learning_rate": 3.9355040701314964e-05,
922
+ "loss": 0.139,
923
+ "step": 1180
924
+ },
925
+ {
926
+ "epoch": 0.32214401732539255,
927
+ "grad_norm": 1.03125,
928
+ "learning_rate": 3.919849718221666e-05,
929
+ "loss": 0.1405,
930
+ "step": 1190
931
+ },
932
+ {
933
+ "epoch": 0.32485110990795885,
934
+ "grad_norm": 0.75390625,
935
+ "learning_rate": 3.9041953663118345e-05,
936
+ "loss": 0.1359,
937
+ "step": 1200
938
+ },
939
+ {
940
+ "epoch": 0.32485110990795885,
941
+ "eval_loss": 0.13801878690719604,
942
+ "eval_runtime": 105.4772,
943
+ "eval_samples_per_second": 4.854,
944
+ "eval_steps_per_second": 0.152,
945
+ "step": 1200
946
+ },
947
+ {
948
+ "epoch": 0.3275582024905252,
949
+ "grad_norm": 0.9609375,
950
+ "learning_rate": 3.888541014402004e-05,
951
+ "loss": 0.0934,
952
+ "step": 1210
953
+ },
954
+ {
955
+ "epoch": 0.3302652950730915,
956
+ "grad_norm": 1.9921875,
957
+ "learning_rate": 3.8728866624921726e-05,
958
+ "loss": 0.1272,
959
+ "step": 1220
960
+ },
961
+ {
962
+ "epoch": 0.33297238765565784,
963
+ "grad_norm": 2.734375,
964
+ "learning_rate": 3.8572323105823424e-05,
965
+ "loss": 0.1099,
966
+ "step": 1230
967
+ },
968
+ {
969
+ "epoch": 0.33567948023822414,
970
+ "grad_norm": 1.1171875,
971
+ "learning_rate": 3.841577958672511e-05,
972
+ "loss": 0.1019,
973
+ "step": 1240
974
+ },
975
+ {
976
+ "epoch": 0.3383865728207905,
977
+ "grad_norm": 1.1171875,
978
+ "learning_rate": 3.8259236067626805e-05,
979
+ "loss": 0.1503,
980
+ "step": 1250
981
+ },
982
+ {
983
+ "epoch": 0.3410936654033568,
984
+ "grad_norm": 0.490234375,
985
+ "learning_rate": 3.810269254852849e-05,
986
+ "loss": 0.0857,
987
+ "step": 1260
988
+ },
989
+ {
990
+ "epoch": 0.34380075798592313,
991
+ "grad_norm": 0.69140625,
992
+ "learning_rate": 3.794614902943019e-05,
993
+ "loss": 0.0949,
994
+ "step": 1270
995
+ },
996
+ {
997
+ "epoch": 0.34650785056848943,
998
+ "grad_norm": 2.015625,
999
+ "learning_rate": 3.778960551033187e-05,
1000
+ "loss": 0.1242,
1001
+ "step": 1280
1002
+ },
1003
+ {
1004
+ "epoch": 0.3492149431510558,
1005
+ "grad_norm": 0.5546875,
1006
+ "learning_rate": 3.763306199123356e-05,
1007
+ "loss": 0.1041,
1008
+ "step": 1290
1009
+ },
1010
+ {
1011
+ "epoch": 0.3519220357336221,
1012
+ "grad_norm": 1.5234375,
1013
+ "learning_rate": 3.747651847213526e-05,
1014
+ "loss": 0.0727,
1015
+ "step": 1300
1016
+ },
1017
+ {
1018
+ "epoch": 0.3519220357336221,
1019
+ "eval_loss": 0.13924407958984375,
1020
+ "eval_runtime": 103.9612,
1021
+ "eval_samples_per_second": 4.925,
1022
+ "eval_steps_per_second": 0.154,
1023
+ "step": 1300
1024
+ },
1025
+ {
1026
+ "epoch": 0.3546291283161884,
1027
+ "grad_norm": 1.0625,
1028
+ "learning_rate": 3.731997495303694e-05,
1029
+ "loss": 0.0898,
1030
+ "step": 1310
1031
+ },
1032
+ {
1033
+ "epoch": 0.3573362208987547,
1034
+ "grad_norm": 2.609375,
1035
+ "learning_rate": 3.716343143393864e-05,
1036
+ "loss": 0.1697,
1037
+ "step": 1320
1038
+ },
1039
+ {
1040
+ "epoch": 0.36004331348132107,
1041
+ "grad_norm": 1.1015625,
1042
+ "learning_rate": 3.7006887914840324e-05,
1043
+ "loss": 0.1111,
1044
+ "step": 1330
1045
+ },
1046
+ {
1047
+ "epoch": 0.36275040606388737,
1048
+ "grad_norm": 0.44921875,
1049
+ "learning_rate": 3.685034439574202e-05,
1050
+ "loss": 0.1194,
1051
+ "step": 1340
1052
+ },
1053
+ {
1054
+ "epoch": 0.3654574986464537,
1055
+ "grad_norm": 1.421875,
1056
+ "learning_rate": 3.6693800876643706e-05,
1057
+ "loss": 0.1019,
1058
+ "step": 1350
1059
+ },
1060
+ {
1061
+ "epoch": 0.36816459122902,
1062
+ "grad_norm": 1.25,
1063
+ "learning_rate": 3.65372573575454e-05,
1064
+ "loss": 0.0984,
1065
+ "step": 1360
1066
+ },
1067
+ {
1068
+ "epoch": 0.37087168381158636,
1069
+ "grad_norm": 0.86328125,
1070
+ "learning_rate": 3.638071383844709e-05,
1071
+ "loss": 0.1298,
1072
+ "step": 1370
1073
+ },
1074
+ {
1075
+ "epoch": 0.37357877639415266,
1076
+ "grad_norm": 1.6484375,
1077
+ "learning_rate": 3.6224170319348784e-05,
1078
+ "loss": 0.1242,
1079
+ "step": 1380
1080
+ },
1081
+ {
1082
+ "epoch": 0.376285868976719,
1083
+ "grad_norm": 0.72265625,
1084
+ "learning_rate": 3.606762680025047e-05,
1085
+ "loss": 0.094,
1086
+ "step": 1390
1087
+ },
1088
+ {
1089
+ "epoch": 0.3789929615592853,
1090
+ "grad_norm": 1.2421875,
1091
+ "learning_rate": 3.5911083281152166e-05,
1092
+ "loss": 0.1286,
1093
+ "step": 1400
1094
+ },
1095
+ {
1096
+ "epoch": 0.3789929615592853,
1097
+ "eval_loss": 0.13358080387115479,
1098
+ "eval_runtime": 120.0689,
1099
+ "eval_samples_per_second": 4.264,
1100
+ "eval_steps_per_second": 0.133,
1101
+ "step": 1400
1102
+ },
1103
+ {
1104
+ "epoch": 0.38170005414185165,
1105
+ "grad_norm": 2.703125,
1106
+ "learning_rate": 3.575453976205385e-05,
1107
+ "loss": 0.1594,
1108
+ "step": 1410
1109
+ },
1110
+ {
1111
+ "epoch": 0.38440714672441795,
1112
+ "grad_norm": 2.0,
1113
+ "learning_rate": 3.559799624295555e-05,
1114
+ "loss": 0.0771,
1115
+ "step": 1420
1116
+ },
1117
+ {
1118
+ "epoch": 0.3871142393069843,
1119
+ "grad_norm": 2.328125,
1120
+ "learning_rate": 3.544145272385723e-05,
1121
+ "loss": 0.1485,
1122
+ "step": 1430
1123
+ },
1124
+ {
1125
+ "epoch": 0.38982133188955065,
1126
+ "grad_norm": 0.30859375,
1127
+ "learning_rate": 3.528490920475893e-05,
1128
+ "loss": 0.0938,
1129
+ "step": 1440
1130
+ },
1131
+ {
1132
+ "epoch": 0.39252842447211694,
1133
+ "grad_norm": 0.91015625,
1134
+ "learning_rate": 3.512836568566061e-05,
1135
+ "loss": 0.1372,
1136
+ "step": 1450
1137
+ },
1138
+ {
1139
+ "epoch": 0.3952355170546833,
1140
+ "grad_norm": 0.6953125,
1141
+ "learning_rate": 3.497182216656231e-05,
1142
+ "loss": 0.1033,
1143
+ "step": 1460
1144
+ },
1145
+ {
1146
+ "epoch": 0.3979426096372496,
1147
+ "grad_norm": 1.0546875,
1148
+ "learning_rate": 3.4815278647463994e-05,
1149
+ "loss": 0.1345,
1150
+ "step": 1470
1151
+ },
1152
+ {
1153
+ "epoch": 0.40064970221981594,
1154
+ "grad_norm": 0.91015625,
1155
+ "learning_rate": 3.4658735128365685e-05,
1156
+ "loss": 0.103,
1157
+ "step": 1480
1158
+ },
1159
+ {
1160
+ "epoch": 0.40335679480238223,
1161
+ "grad_norm": 1.125,
1162
+ "learning_rate": 3.4502191609267375e-05,
1163
+ "loss": 0.083,
1164
+ "step": 1490
1165
+ },
1166
+ {
1167
+ "epoch": 0.4060638873849486,
1168
+ "grad_norm": 0.9296875,
1169
+ "learning_rate": 3.4345648090169066e-05,
1170
+ "loss": 0.1024,
1171
+ "step": 1500
1172
+ },
1173
+ {
1174
+ "epoch": 0.4060638873849486,
1175
+ "eval_loss": 0.13767841458320618,
1176
+ "eval_runtime": 104.1987,
1177
+ "eval_samples_per_second": 4.914,
1178
+ "eval_steps_per_second": 0.154,
1179
+ "step": 1500
1180
+ },
1181
+ {
1182
+ "epoch": 0.4087709799675149,
1183
+ "grad_norm": 1.71875,
1184
+ "learning_rate": 3.418910457107076e-05,
1185
+ "loss": 0.1319,
1186
+ "step": 1510
1187
+ },
1188
+ {
1189
+ "epoch": 0.41147807255008123,
1190
+ "grad_norm": 1.0703125,
1191
+ "learning_rate": 3.403256105197245e-05,
1192
+ "loss": 0.1152,
1193
+ "step": 1520
1194
+ },
1195
+ {
1196
+ "epoch": 0.4141851651326475,
1197
+ "grad_norm": 1.3125,
1198
+ "learning_rate": 3.387601753287414e-05,
1199
+ "loss": 0.1072,
1200
+ "step": 1530
1201
+ },
1202
+ {
1203
+ "epoch": 0.4168922577152139,
1204
+ "grad_norm": 2.171875,
1205
+ "learning_rate": 3.371947401377583e-05,
1206
+ "loss": 0.1408,
1207
+ "step": 1540
1208
+ },
1209
+ {
1210
+ "epoch": 0.41959935029778017,
1211
+ "grad_norm": 1.53125,
1212
+ "learning_rate": 3.3562930494677526e-05,
1213
+ "loss": 0.1125,
1214
+ "step": 1550
1215
+ },
1216
+ {
1217
+ "epoch": 0.4223064428803465,
1218
+ "grad_norm": 1.578125,
1219
+ "learning_rate": 3.340638697557921e-05,
1220
+ "loss": 0.1144,
1221
+ "step": 1560
1222
+ },
1223
+ {
1224
+ "epoch": 0.4250135354629128,
1225
+ "grad_norm": 1.6015625,
1226
+ "learning_rate": 3.324984345648091e-05,
1227
+ "loss": 0.1092,
1228
+ "step": 1570
1229
+ },
1230
+ {
1231
+ "epoch": 0.42772062804547917,
1232
+ "grad_norm": 0.408203125,
1233
+ "learning_rate": 3.309329993738259e-05,
1234
+ "loss": 0.1271,
1235
+ "step": 1580
1236
+ },
1237
+ {
1238
+ "epoch": 0.43042772062804546,
1239
+ "grad_norm": 0.72265625,
1240
+ "learning_rate": 3.293675641828429e-05,
1241
+ "loss": 0.0882,
1242
+ "step": 1590
1243
+ },
1244
+ {
1245
+ "epoch": 0.4331348132106118,
1246
+ "grad_norm": 0.296875,
1247
+ "learning_rate": 3.278021289918597e-05,
1248
+ "loss": 0.1141,
1249
+ "step": 1600
1250
+ },
1251
+ {
1252
+ "epoch": 0.4331348132106118,
1253
+ "eval_loss": 0.13831546902656555,
1254
+ "eval_runtime": 104.916,
1255
+ "eval_samples_per_second": 4.88,
1256
+ "eval_steps_per_second": 0.153,
1257
+ "step": 1600
1258
+ },
1259
+ {
1260
+ "epoch": 0.4358419057931781,
1261
+ "grad_norm": 0.462890625,
1262
+ "learning_rate": 3.262366938008767e-05,
1263
+ "loss": 0.0874,
1264
+ "step": 1610
1265
+ },
1266
+ {
1267
+ "epoch": 0.43854899837574446,
1268
+ "grad_norm": 1.921875,
1269
+ "learning_rate": 3.2467125860989355e-05,
1270
+ "loss": 0.1288,
1271
+ "step": 1620
1272
+ },
1273
+ {
1274
+ "epoch": 0.44125609095831075,
1275
+ "grad_norm": 1.3359375,
1276
+ "learning_rate": 3.231058234189105e-05,
1277
+ "loss": 0.1171,
1278
+ "step": 1630
1279
+ },
1280
+ {
1281
+ "epoch": 0.4439631835408771,
1282
+ "grad_norm": 0.447265625,
1283
+ "learning_rate": 3.2154038822792736e-05,
1284
+ "loss": 0.0845,
1285
+ "step": 1640
1286
+ },
1287
+ {
1288
+ "epoch": 0.4466702761234434,
1289
+ "grad_norm": 1.703125,
1290
+ "learning_rate": 3.1997495303694433e-05,
1291
+ "loss": 0.1052,
1292
+ "step": 1650
1293
+ },
1294
+ {
1295
+ "epoch": 0.44937736870600975,
1296
+ "grad_norm": 1.484375,
1297
+ "learning_rate": 3.184095178459612e-05,
1298
+ "loss": 0.1187,
1299
+ "step": 1660
1300
+ },
1301
+ {
1302
+ "epoch": 0.45208446128857604,
1303
+ "grad_norm": 2.8125,
1304
+ "learning_rate": 3.168440826549781e-05,
1305
+ "loss": 0.1051,
1306
+ "step": 1670
1307
+ },
1308
+ {
1309
+ "epoch": 0.4547915538711424,
1310
+ "grad_norm": 1.9765625,
1311
+ "learning_rate": 3.15278647463995e-05,
1312
+ "loss": 0.1415,
1313
+ "step": 1680
1314
+ },
1315
+ {
1316
+ "epoch": 0.4574986464537087,
1317
+ "grad_norm": 1.3828125,
1318
+ "learning_rate": 3.137132122730119e-05,
1319
+ "loss": 0.1313,
1320
+ "step": 1690
1321
+ },
1322
+ {
1323
+ "epoch": 0.46020573903627504,
1324
+ "grad_norm": 0.8359375,
1325
+ "learning_rate": 3.121477770820288e-05,
1326
+ "loss": 0.1129,
1327
+ "step": 1700
1328
+ },
1329
+ {
1330
+ "epoch": 0.46020573903627504,
1331
+ "eval_loss": 0.13771241903305054,
1332
+ "eval_runtime": 104.8503,
1333
+ "eval_samples_per_second": 4.883,
1334
+ "eval_steps_per_second": 0.153,
1335
+ "step": 1700
1336
+ },
1337
+ {
1338
+ "epoch": 0.4629128316188414,
1339
+ "grad_norm": 1.8359375,
1340
+ "learning_rate": 3.105823418910457e-05,
1341
+ "loss": 0.1618,
1342
+ "step": 1710
1343
+ },
1344
+ {
1345
+ "epoch": 0.4656199242014077,
1346
+ "grad_norm": 0.53515625,
1347
+ "learning_rate": 3.090169067000626e-05,
1348
+ "loss": 0.1025,
1349
+ "step": 1720
1350
+ },
1351
+ {
1352
+ "epoch": 0.46832701678397404,
1353
+ "grad_norm": 4.0625,
1354
+ "learning_rate": 3.074514715090795e-05,
1355
+ "loss": 0.1253,
1356
+ "step": 1730
1357
+ },
1358
+ {
1359
+ "epoch": 0.47103410936654033,
1360
+ "grad_norm": 2.25,
1361
+ "learning_rate": 3.058860363180964e-05,
1362
+ "loss": 0.1098,
1363
+ "step": 1740
1364
+ },
1365
+ {
1366
+ "epoch": 0.4737412019491067,
1367
+ "grad_norm": 1.484375,
1368
+ "learning_rate": 3.0432060112711337e-05,
1369
+ "loss": 0.0868,
1370
+ "step": 1750
1371
+ },
1372
+ {
1373
+ "epoch": 0.476448294531673,
1374
+ "grad_norm": 1.9375,
1375
+ "learning_rate": 3.0275516593613024e-05,
1376
+ "loss": 0.1427,
1377
+ "step": 1760
1378
+ },
1379
+ {
1380
+ "epoch": 0.47915538711423933,
1381
+ "grad_norm": 2.546875,
1382
+ "learning_rate": 3.011897307451472e-05,
1383
+ "loss": 0.1272,
1384
+ "step": 1770
1385
+ },
1386
+ {
1387
+ "epoch": 0.4818624796968056,
1388
+ "grad_norm": 2.703125,
1389
+ "learning_rate": 2.9962429555416406e-05,
1390
+ "loss": 0.16,
1391
+ "step": 1780
1392
+ },
1393
+ {
1394
+ "epoch": 0.484569572279372,
1395
+ "grad_norm": 1.2578125,
1396
+ "learning_rate": 2.9805886036318097e-05,
1397
+ "loss": 0.1079,
1398
+ "step": 1790
1399
+ },
1400
+ {
1401
+ "epoch": 0.48727666486193827,
1402
+ "grad_norm": 0.75,
1403
+ "learning_rate": 2.9649342517219787e-05,
1404
+ "loss": 0.0907,
1405
+ "step": 1800
1406
+ },
1407
+ {
1408
+ "epoch": 0.48727666486193827,
1409
+ "eval_loss": 0.14291706681251526,
1410
+ "eval_runtime": 105.4966,
1411
+ "eval_samples_per_second": 4.853,
1412
+ "eval_steps_per_second": 0.152,
1413
+ "step": 1800
1414
+ },
1415
+ {
1416
+ "epoch": 0.4899837574445046,
1417
+ "grad_norm": 0.94140625,
1418
+ "learning_rate": 2.9492798998121478e-05,
1419
+ "loss": 0.1137,
1420
+ "step": 1810
1421
+ },
1422
+ {
1423
+ "epoch": 0.4926908500270709,
1424
+ "grad_norm": 1.796875,
1425
+ "learning_rate": 2.9336255479023172e-05,
1426
+ "loss": 0.114,
1427
+ "step": 1820
1428
+ },
1429
+ {
1430
+ "epoch": 0.49539794260963727,
1431
+ "grad_norm": 1.4296875,
1432
+ "learning_rate": 2.917971195992486e-05,
1433
+ "loss": 0.1263,
1434
+ "step": 1830
1435
+ },
1436
+ {
1437
+ "epoch": 0.49810503519220356,
1438
+ "grad_norm": 2.171875,
1439
+ "learning_rate": 2.9023168440826553e-05,
1440
+ "loss": 0.1095,
1441
+ "step": 1840
1442
+ },
1443
+ {
1444
+ "epoch": 0.5008121277747699,
1445
+ "grad_norm": 1.2734375,
1446
+ "learning_rate": 2.886662492172824e-05,
1447
+ "loss": 0.1609,
1448
+ "step": 1850
1449
+ },
1450
+ {
1451
+ "epoch": 0.5035192203573362,
1452
+ "grad_norm": 1.84375,
1453
+ "learning_rate": 2.8710081402629935e-05,
1454
+ "loss": 0.1079,
1455
+ "step": 1860
1456
+ },
1457
+ {
1458
+ "epoch": 0.5062263129399025,
1459
+ "grad_norm": 0.328125,
1460
+ "learning_rate": 2.8553537883531622e-05,
1461
+ "loss": 0.1114,
1462
+ "step": 1870
1463
+ },
1464
+ {
1465
+ "epoch": 0.5089334055224689,
1466
+ "grad_norm": 1.9140625,
1467
+ "learning_rate": 2.8396994364433316e-05,
1468
+ "loss": 0.1163,
1469
+ "step": 1880
1470
+ },
1471
+ {
1472
+ "epoch": 0.5116404981050352,
1473
+ "grad_norm": 0.68359375,
1474
+ "learning_rate": 2.8240450845335004e-05,
1475
+ "loss": 0.105,
1476
+ "step": 1890
1477
+ },
1478
+ {
1479
+ "epoch": 0.5143475906876015,
1480
+ "grad_norm": 0.76171875,
1481
+ "learning_rate": 2.8083907326236698e-05,
1482
+ "loss": 0.0931,
1483
+ "step": 1900
1484
+ },
1485
+ {
1486
+ "epoch": 0.5143475906876015,
1487
+ "eval_loss": 0.1378411054611206,
1488
+ "eval_runtime": 122.4527,
1489
+ "eval_samples_per_second": 4.181,
1490
+ "eval_steps_per_second": 0.131,
1491
+ "step": 1900
1492
+ },
1493
+ {
1494
+ "epoch": 0.5170546832701678,
1495
+ "grad_norm": 2.125,
1496
+ "learning_rate": 2.7927363807138385e-05,
1497
+ "loss": 0.1137,
1498
+ "step": 1910
1499
+ },
1500
+ {
1501
+ "epoch": 0.5197617758527342,
1502
+ "grad_norm": 1.265625,
1503
+ "learning_rate": 2.777082028804008e-05,
1504
+ "loss": 0.1127,
1505
+ "step": 1920
1506
+ },
1507
+ {
1508
+ "epoch": 0.5224688684353005,
1509
+ "grad_norm": 1.7734375,
1510
+ "learning_rate": 2.7614276768941766e-05,
1511
+ "loss": 0.1082,
1512
+ "step": 1930
1513
+ },
1514
+ {
1515
+ "epoch": 0.5251759610178668,
1516
+ "grad_norm": 1.6640625,
1517
+ "learning_rate": 2.745773324984346e-05,
1518
+ "loss": 0.1147,
1519
+ "step": 1940
1520
+ },
1521
+ {
1522
+ "epoch": 0.5278830536004331,
1523
+ "grad_norm": 1.84375,
1524
+ "learning_rate": 2.7301189730745148e-05,
1525
+ "loss": 0.1027,
1526
+ "step": 1950
1527
+ },
1528
+ {
1529
+ "epoch": 0.5305901461829995,
1530
+ "grad_norm": 0.86328125,
1531
+ "learning_rate": 2.7144646211646842e-05,
1532
+ "loss": 0.108,
1533
+ "step": 1960
1534
+ },
1535
+ {
1536
+ "epoch": 0.5332972387655658,
1537
+ "grad_norm": 0.59375,
1538
+ "learning_rate": 2.698810269254853e-05,
1539
+ "loss": 0.0917,
1540
+ "step": 1970
1541
+ },
1542
+ {
1543
+ "epoch": 0.5360043313481321,
1544
+ "grad_norm": 0.515625,
1545
+ "learning_rate": 2.683155917345022e-05,
1546
+ "loss": 0.078,
1547
+ "step": 1980
1548
+ },
1549
+ {
1550
+ "epoch": 0.5387114239306985,
1551
+ "grad_norm": 1.1796875,
1552
+ "learning_rate": 2.667501565435191e-05,
1553
+ "loss": 0.0845,
1554
+ "step": 1990
1555
+ },
1556
+ {
1557
+ "epoch": 0.5414185165132648,
1558
+ "grad_norm": 1.7890625,
1559
+ "learning_rate": 2.65184721352536e-05,
1560
+ "loss": 0.0888,
1561
+ "step": 2000
1562
+ },
1563
+ {
1564
+ "epoch": 0.5414185165132648,
1565
+ "eval_loss": 0.14294707775115967,
1566
+ "eval_runtime": 103.5797,
1567
+ "eval_samples_per_second": 4.943,
1568
+ "eval_steps_per_second": 0.154,
1569
+ "step": 2000
1570
+ },
1571
+ {
1572
+ "epoch": 0.5441256090958311,
1573
+ "grad_norm": 2.234375,
1574
+ "learning_rate": 2.636192861615529e-05,
1575
+ "loss": 0.099,
1576
+ "step": 2010
1577
+ },
1578
+ {
1579
+ "epoch": 0.5468327016783974,
1580
+ "grad_norm": 1.2421875,
1581
+ "learning_rate": 2.6205385097056983e-05,
1582
+ "loss": 0.1048,
1583
+ "step": 2020
1584
+ },
1585
+ {
1586
+ "epoch": 0.5495397942609638,
1587
+ "grad_norm": 1.984375,
1588
+ "learning_rate": 2.604884157795867e-05,
1589
+ "loss": 0.1117,
1590
+ "step": 2030
1591
+ },
1592
+ {
1593
+ "epoch": 0.5522468868435301,
1594
+ "grad_norm": 1.6328125,
1595
+ "learning_rate": 2.5892298058860364e-05,
1596
+ "loss": 0.1291,
1597
+ "step": 2040
1598
+ },
1599
+ {
1600
+ "epoch": 0.5549539794260964,
1601
+ "grad_norm": 1.4765625,
1602
+ "learning_rate": 2.573575453976205e-05,
1603
+ "loss": 0.1529,
1604
+ "step": 2050
1605
+ },
1606
+ {
1607
+ "epoch": 0.5576610720086627,
1608
+ "grad_norm": 1.203125,
1609
+ "learning_rate": 2.5579211020663746e-05,
1610
+ "loss": 0.0559,
1611
+ "step": 2060
1612
+ },
1613
+ {
1614
+ "epoch": 0.5603681645912291,
1615
+ "grad_norm": 1.40625,
1616
+ "learning_rate": 2.5422667501565433e-05,
1617
+ "loss": 0.0992,
1618
+ "step": 2070
1619
+ },
1620
+ {
1621
+ "epoch": 0.5630752571737954,
1622
+ "grad_norm": 0.7578125,
1623
+ "learning_rate": 2.5266123982467127e-05,
1624
+ "loss": 0.0876,
1625
+ "step": 2080
1626
+ },
1627
+ {
1628
+ "epoch": 0.5657823497563617,
1629
+ "grad_norm": 1.6328125,
1630
+ "learning_rate": 2.510958046336882e-05,
1631
+ "loss": 0.1442,
1632
+ "step": 2090
1633
+ },
1634
+ {
1635
+ "epoch": 0.568489442338928,
1636
+ "grad_norm": 0.734375,
1637
+ "learning_rate": 2.495303694427051e-05,
1638
+ "loss": 0.073,
1639
+ "step": 2100
1640
+ },
1641
+ {
1642
+ "epoch": 0.568489442338928,
1643
+ "eval_loss": 0.1371915191411972,
1644
+ "eval_runtime": 102.7955,
1645
+ "eval_samples_per_second": 4.981,
1646
+ "eval_steps_per_second": 0.156,
1647
+ "step": 2100
1648
+ },
1649
+ {
1650
+ "epoch": 0.5711965349214944,
1651
+ "grad_norm": 1.4453125,
1652
+ "learning_rate": 2.47964934251722e-05,
1653
+ "loss": 0.0818,
1654
+ "step": 2110
1655
+ },
1656
+ {
1657
+ "epoch": 0.5739036275040607,
1658
+ "grad_norm": 1.0,
1659
+ "learning_rate": 2.463994990607389e-05,
1660
+ "loss": 0.0785,
1661
+ "step": 2120
1662
+ },
1663
+ {
1664
+ "epoch": 0.576610720086627,
1665
+ "grad_norm": 2.09375,
1666
+ "learning_rate": 2.448340638697558e-05,
1667
+ "loss": 0.0966,
1668
+ "step": 2130
1669
+ },
1670
+ {
1671
+ "epoch": 0.5793178126691932,
1672
+ "grad_norm": 1.3828125,
1673
+ "learning_rate": 2.432686286787727e-05,
1674
+ "loss": 0.0805,
1675
+ "step": 2140
1676
+ },
1677
+ {
1678
+ "epoch": 0.5820249052517596,
1679
+ "grad_norm": 0.9375,
1680
+ "learning_rate": 2.4170319348778962e-05,
1681
+ "loss": 0.1032,
1682
+ "step": 2150
1683
+ },
1684
+ {
1685
+ "epoch": 0.584731997834326,
1686
+ "grad_norm": 1.9765625,
1687
+ "learning_rate": 2.4013775829680653e-05,
1688
+ "loss": 0.1314,
1689
+ "step": 2160
1690
+ },
1691
+ {
1692
+ "epoch": 0.5874390904168922,
1693
+ "grad_norm": 1.5078125,
1694
+ "learning_rate": 2.3857232310582343e-05,
1695
+ "loss": 0.1026,
1696
+ "step": 2170
1697
+ },
1698
+ {
1699
+ "epoch": 0.5901461829994585,
1700
+ "grad_norm": 0.58984375,
1701
+ "learning_rate": 2.3700688791484034e-05,
1702
+ "loss": 0.0858,
1703
+ "step": 2180
1704
+ },
1705
+ {
1706
+ "epoch": 0.5928532755820249,
1707
+ "grad_norm": 0.51171875,
1708
+ "learning_rate": 2.3544145272385725e-05,
1709
+ "loss": 0.069,
1710
+ "step": 2190
1711
+ },
1712
+ {
1713
+ "epoch": 0.5955603681645912,
1714
+ "grad_norm": 0.94140625,
1715
+ "learning_rate": 2.3387601753287412e-05,
1716
+ "loss": 0.1238,
1717
+ "step": 2200
1718
+ },
1719
+ {
1720
+ "epoch": 0.5955603681645912,
1721
+ "eval_loss": 0.1365373730659485,
1722
+ "eval_runtime": 103.5823,
1723
+ "eval_samples_per_second": 4.943,
1724
+ "eval_steps_per_second": 0.154,
1725
+ "step": 2200
1726
+ },
1727
+ {
1728
+ "epoch": 0.5982674607471575,
1729
+ "grad_norm": 1.0390625,
1730
+ "learning_rate": 2.3231058234189106e-05,
1731
+ "loss": 0.1196,
1732
+ "step": 2210
1733
+ },
1734
+ {
1735
+ "epoch": 0.6009745533297238,
1736
+ "grad_norm": 1.1875,
1737
+ "learning_rate": 2.3074514715090797e-05,
1738
+ "loss": 0.111,
1739
+ "step": 2220
1740
+ },
1741
+ {
1742
+ "epoch": 0.6036816459122902,
1743
+ "grad_norm": 0.5546875,
1744
+ "learning_rate": 2.2917971195992488e-05,
1745
+ "loss": 0.0875,
1746
+ "step": 2230
1747
+ },
1748
+ {
1749
+ "epoch": 0.6063887384948565,
1750
+ "grad_norm": 1.2421875,
1751
+ "learning_rate": 2.2761427676894178e-05,
1752
+ "loss": 0.1029,
1753
+ "step": 2240
1754
+ },
1755
+ {
1756
+ "epoch": 0.6090958310774228,
1757
+ "grad_norm": 1.3671875,
1758
+ "learning_rate": 2.260488415779587e-05,
1759
+ "loss": 0.0933,
1760
+ "step": 2250
1761
+ },
1762
+ {
1763
+ "epoch": 0.6118029236599891,
1764
+ "grad_norm": 1.140625,
1765
+ "learning_rate": 2.244834063869756e-05,
1766
+ "loss": 0.1009,
1767
+ "step": 2260
1768
+ },
1769
+ {
1770
+ "epoch": 0.6145100162425555,
1771
+ "grad_norm": 2.234375,
1772
+ "learning_rate": 2.229179711959925e-05,
1773
+ "loss": 0.135,
1774
+ "step": 2270
1775
+ },
1776
+ {
1777
+ "epoch": 0.6172171088251218,
1778
+ "grad_norm": 1.671875,
1779
+ "learning_rate": 2.213525360050094e-05,
1780
+ "loss": 0.0773,
1781
+ "step": 2280
1782
+ },
1783
+ {
1784
+ "epoch": 0.6199242014076881,
1785
+ "grad_norm": 0.9375,
1786
+ "learning_rate": 2.1978710081402632e-05,
1787
+ "loss": 0.0732,
1788
+ "step": 2290
1789
+ },
1790
+ {
1791
+ "epoch": 0.6226312939902545,
1792
+ "grad_norm": 2.0625,
1793
+ "learning_rate": 2.1822166562304323e-05,
1794
+ "loss": 0.0851,
1795
+ "step": 2300
1796
+ },
1797
+ {
1798
+ "epoch": 0.6226312939902545,
1799
+ "eval_loss": 0.1353287696838379,
1800
+ "eval_runtime": 102.3815,
1801
+ "eval_samples_per_second": 5.001,
1802
+ "eval_steps_per_second": 0.156,
1803
+ "step": 2300
1804
+ },
1805
+ {
1806
+ "epoch": 0.6253383865728208,
1807
+ "grad_norm": 0.71484375,
1808
+ "learning_rate": 2.1665623043206013e-05,
1809
+ "loss": 0.0633,
1810
+ "step": 2310
1811
+ },
1812
+ {
1813
+ "epoch": 0.6280454791553871,
1814
+ "grad_norm": 1.328125,
1815
+ "learning_rate": 2.1509079524107704e-05,
1816
+ "loss": 0.0825,
1817
+ "step": 2320
1818
+ },
1819
+ {
1820
+ "epoch": 0.6307525717379534,
1821
+ "grad_norm": 0.515625,
1822
+ "learning_rate": 2.1352536005009395e-05,
1823
+ "loss": 0.1016,
1824
+ "step": 2330
1825
+ },
1826
+ {
1827
+ "epoch": 0.6334596643205198,
1828
+ "grad_norm": 2.40625,
1829
+ "learning_rate": 2.1195992485911085e-05,
1830
+ "loss": 0.0942,
1831
+ "step": 2340
1832
+ },
1833
+ {
1834
+ "epoch": 0.6361667569030861,
1835
+ "grad_norm": 1.75,
1836
+ "learning_rate": 2.1039448966812776e-05,
1837
+ "loss": 0.086,
1838
+ "step": 2350
1839
+ },
1840
+ {
1841
+ "epoch": 0.6388738494856524,
1842
+ "grad_norm": 1.328125,
1843
+ "learning_rate": 2.0882905447714467e-05,
1844
+ "loss": 0.1095,
1845
+ "step": 2360
1846
+ },
1847
+ {
1848
+ "epoch": 0.6415809420682187,
1849
+ "grad_norm": 0.63671875,
1850
+ "learning_rate": 2.0726361928616157e-05,
1851
+ "loss": 0.0711,
1852
+ "step": 2370
1853
+ },
1854
+ {
1855
+ "epoch": 0.6442880346507851,
1856
+ "grad_norm": 1.3515625,
1857
+ "learning_rate": 2.0569818409517845e-05,
1858
+ "loss": 0.0677,
1859
+ "step": 2380
1860
+ },
1861
+ {
1862
+ "epoch": 0.6469951272333514,
1863
+ "grad_norm": 2.046875,
1864
+ "learning_rate": 2.0413274890419535e-05,
1865
+ "loss": 0.0892,
1866
+ "step": 2390
1867
+ },
1868
+ {
1869
+ "epoch": 0.6497022198159177,
1870
+ "grad_norm": 1.640625,
1871
+ "learning_rate": 2.0256731371321226e-05,
1872
+ "loss": 0.0955,
1873
+ "step": 2400
1874
+ },
1875
+ {
1876
+ "epoch": 0.6497022198159177,
1877
+ "eval_loss": 0.13519829511642456,
1878
+ "eval_runtime": 121.253,
1879
+ "eval_samples_per_second": 4.223,
1880
+ "eval_steps_per_second": 0.132,
1881
+ "step": 2400
1882
+ },
1883
+ {
1884
+ "epoch": 0.652409312398484,
1885
+ "grad_norm": 1.265625,
1886
+ "learning_rate": 2.0100187852222917e-05,
1887
+ "loss": 0.1012,
1888
+ "step": 2410
1889
+ },
1890
+ {
1891
+ "epoch": 0.6551164049810504,
1892
+ "grad_norm": 0.984375,
1893
+ "learning_rate": 1.9943644333124608e-05,
1894
+ "loss": 0.1008,
1895
+ "step": 2420
1896
+ },
1897
+ {
1898
+ "epoch": 0.6578234975636167,
1899
+ "grad_norm": 0.9453125,
1900
+ "learning_rate": 1.9787100814026298e-05,
1901
+ "loss": 0.1026,
1902
+ "step": 2430
1903
+ },
1904
+ {
1905
+ "epoch": 0.660530590146183,
1906
+ "grad_norm": 0.95703125,
1907
+ "learning_rate": 1.963055729492799e-05,
1908
+ "loss": 0.1021,
1909
+ "step": 2440
1910
+ },
1911
+ {
1912
+ "epoch": 0.6632376827287493,
1913
+ "grad_norm": 1.4375,
1914
+ "learning_rate": 1.947401377582968e-05,
1915
+ "loss": 0.1098,
1916
+ "step": 2450
1917
+ },
1918
+ {
1919
+ "epoch": 0.6659447753113157,
1920
+ "grad_norm": 0.375,
1921
+ "learning_rate": 1.931747025673137e-05,
1922
+ "loss": 0.0828,
1923
+ "step": 2460
1924
+ },
1925
+ {
1926
+ "epoch": 0.668651867893882,
1927
+ "grad_norm": 0.94140625,
1928
+ "learning_rate": 1.9160926737633064e-05,
1929
+ "loss": 0.0612,
1930
+ "step": 2470
1931
+ },
1932
+ {
1933
+ "epoch": 0.6713589604764483,
1934
+ "grad_norm": 3.21875,
1935
+ "learning_rate": 1.9004383218534755e-05,
1936
+ "loss": 0.1367,
1937
+ "step": 2480
1938
+ },
1939
+ {
1940
+ "epoch": 0.6740660530590146,
1941
+ "grad_norm": 1.1796875,
1942
+ "learning_rate": 1.8847839699436446e-05,
1943
+ "loss": 0.0647,
1944
+ "step": 2490
1945
+ },
1946
+ {
1947
+ "epoch": 0.676773145641581,
1948
+ "grad_norm": 0.5703125,
1949
+ "learning_rate": 1.8691296180338137e-05,
1950
+ "loss": 0.0973,
1951
+ "step": 2500
1952
+ },
1953
+ {
1954
+ "epoch": 0.676773145641581,
1955
+ "eval_loss": 0.13495835661888123,
1956
+ "eval_runtime": 105.5301,
1957
+ "eval_samples_per_second": 4.852,
1958
+ "eval_steps_per_second": 0.152,
1959
+ "step": 2500
1960
+ },
1961
+ {
1962
+ "epoch": 0.6794802382241473,
1963
+ "grad_norm": 1.2890625,
1964
+ "learning_rate": 1.8534752661239827e-05,
1965
+ "loss": 0.0775,
1966
+ "step": 2510
1967
+ },
1968
+ {
1969
+ "epoch": 0.6821873308067136,
1970
+ "grad_norm": 0.96484375,
1971
+ "learning_rate": 1.8378209142141518e-05,
1972
+ "loss": 0.0836,
1973
+ "step": 2520
1974
+ },
1975
+ {
1976
+ "epoch": 0.6848944233892799,
1977
+ "grad_norm": 2.5625,
1978
+ "learning_rate": 1.822166562304321e-05,
1979
+ "loss": 0.1118,
1980
+ "step": 2530
1981
+ },
1982
+ {
1983
+ "epoch": 0.6876015159718463,
1984
+ "grad_norm": 1.1328125,
1985
+ "learning_rate": 1.80651221039449e-05,
1986
+ "loss": 0.0889,
1987
+ "step": 2540
1988
+ },
1989
+ {
1990
+ "epoch": 0.6903086085544126,
1991
+ "grad_norm": 0.71484375,
1992
+ "learning_rate": 1.790857858484659e-05,
1993
+ "loss": 0.0982,
1994
+ "step": 2550
1995
+ },
1996
+ {
1997
+ "epoch": 0.6930157011369789,
1998
+ "grad_norm": 2.125,
1999
+ "learning_rate": 1.775203506574828e-05,
2000
+ "loss": 0.1131,
2001
+ "step": 2560
2002
+ },
2003
+ {
2004
+ "epoch": 0.6957227937195453,
2005
+ "grad_norm": 1.015625,
2006
+ "learning_rate": 1.7595491546649968e-05,
2007
+ "loss": 0.1142,
2008
+ "step": 2570
2009
+ },
2010
+ {
2011
+ "epoch": 0.6984298863021116,
2012
+ "grad_norm": 0.3671875,
2013
+ "learning_rate": 1.743894802755166e-05,
2014
+ "loss": 0.0911,
2015
+ "step": 2580
2016
+ },
2017
+ {
2018
+ "epoch": 0.7011369788846779,
2019
+ "grad_norm": 1.2421875,
2020
+ "learning_rate": 1.728240450845335e-05,
2021
+ "loss": 0.0774,
2022
+ "step": 2590
2023
+ },
2024
+ {
2025
+ "epoch": 0.7038440714672441,
2026
+ "grad_norm": 0.9453125,
2027
+ "learning_rate": 1.712586098935504e-05,
2028
+ "loss": 0.0933,
2029
+ "step": 2600
2030
+ },
2031
+ {
2032
+ "epoch": 0.7038440714672441,
2033
+ "eval_loss": 0.13393962383270264,
2034
+ "eval_runtime": 103.5576,
2035
+ "eval_samples_per_second": 4.944,
2036
+ "eval_steps_per_second": 0.155,
2037
+ "step": 2600
2038
+ },
2039
+ {
2040
+ "epoch": 0.7065511640498106,
2041
+ "grad_norm": 1.234375,
2042
+ "learning_rate": 1.696931747025673e-05,
2043
+ "loss": 0.1073,
2044
+ "step": 2610
2045
+ },
2046
+ {
2047
+ "epoch": 0.7092582566323768,
2048
+ "grad_norm": 2.203125,
2049
+ "learning_rate": 1.681277395115842e-05,
2050
+ "loss": 0.097,
2051
+ "step": 2620
2052
+ },
2053
+ {
2054
+ "epoch": 0.7119653492149431,
2055
+ "grad_norm": 0.94140625,
2056
+ "learning_rate": 1.6656230432060112e-05,
2057
+ "loss": 0.0772,
2058
+ "step": 2630
2059
+ },
2060
+ {
2061
+ "epoch": 0.7146724417975094,
2062
+ "grad_norm": 2.953125,
2063
+ "learning_rate": 1.6499686912961803e-05,
2064
+ "loss": 0.0935,
2065
+ "step": 2640
2066
+ },
2067
+ {
2068
+ "epoch": 0.7173795343800758,
2069
+ "grad_norm": 0.267578125,
2070
+ "learning_rate": 1.6343143393863494e-05,
2071
+ "loss": 0.0924,
2072
+ "step": 2650
2073
+ },
2074
+ {
2075
+ "epoch": 0.7200866269626421,
2076
+ "grad_norm": 1.359375,
2077
+ "learning_rate": 1.6186599874765184e-05,
2078
+ "loss": 0.1057,
2079
+ "step": 2660
2080
+ },
2081
+ {
2082
+ "epoch": 0.7227937195452084,
2083
+ "grad_norm": 1.4296875,
2084
+ "learning_rate": 1.6030056355666875e-05,
2085
+ "loss": 0.0865,
2086
+ "step": 2670
2087
+ },
2088
+ {
2089
+ "epoch": 0.7255008121277747,
2090
+ "grad_norm": 1.09375,
2091
+ "learning_rate": 1.5873512836568566e-05,
2092
+ "loss": 0.0957,
2093
+ "step": 2680
2094
+ },
2095
+ {
2096
+ "epoch": 0.7282079047103411,
2097
+ "grad_norm": 1.09375,
2098
+ "learning_rate": 1.5716969317470257e-05,
2099
+ "loss": 0.107,
2100
+ "step": 2690
2101
+ },
2102
+ {
2103
+ "epoch": 0.7309149972929074,
2104
+ "grad_norm": 0.734375,
2105
+ "learning_rate": 1.5560425798371947e-05,
2106
+ "loss": 0.0976,
2107
+ "step": 2700
2108
+ },
2109
+ {
2110
+ "epoch": 0.7309149972929074,
2111
+ "eval_loss": 0.13289867341518402,
2112
+ "eval_runtime": 104.5232,
2113
+ "eval_samples_per_second": 4.898,
2114
+ "eval_steps_per_second": 0.153,
2115
+ "step": 2700
2116
+ },
2117
+ {
2118
+ "epoch": 0.7336220898754737,
2119
+ "grad_norm": 0.34765625,
2120
+ "learning_rate": 1.5403882279273638e-05,
2121
+ "loss": 0.0918,
2122
+ "step": 2710
2123
+ },
2124
+ {
2125
+ "epoch": 0.73632918245804,
2126
+ "grad_norm": 1.6875,
2127
+ "learning_rate": 1.5247338760175329e-05,
2128
+ "loss": 0.1113,
2129
+ "step": 2720
2130
+ },
2131
+ {
2132
+ "epoch": 0.7390362750406064,
2133
+ "grad_norm": 1.328125,
2134
+ "learning_rate": 1.5090795241077021e-05,
2135
+ "loss": 0.1221,
2136
+ "step": 2730
2137
+ },
2138
+ {
2139
+ "epoch": 0.7417433676231727,
2140
+ "grad_norm": 1.0390625,
2141
+ "learning_rate": 1.4934251721978712e-05,
2142
+ "loss": 0.1236,
2143
+ "step": 2740
2144
+ },
2145
+ {
2146
+ "epoch": 0.744450460205739,
2147
+ "grad_norm": 0.95703125,
2148
+ "learning_rate": 1.4777708202880403e-05,
2149
+ "loss": 0.119,
2150
+ "step": 2750
2151
+ },
2152
+ {
2153
+ "epoch": 0.7471575527883053,
2154
+ "grad_norm": 3.125,
2155
+ "learning_rate": 1.4621164683782093e-05,
2156
+ "loss": 0.1141,
2157
+ "step": 2760
2158
+ },
2159
+ {
2160
+ "epoch": 0.7498646453708717,
2161
+ "grad_norm": 0.94921875,
2162
+ "learning_rate": 1.4464621164683784e-05,
2163
+ "loss": 0.106,
2164
+ "step": 2770
2165
+ },
2166
+ {
2167
+ "epoch": 0.752571737953438,
2168
+ "grad_norm": 1.1640625,
2169
+ "learning_rate": 1.4308077645585475e-05,
2170
+ "loss": 0.0963,
2171
+ "step": 2780
2172
+ },
2173
+ {
2174
+ "epoch": 0.7552788305360043,
2175
+ "grad_norm": 1.234375,
2176
+ "learning_rate": 1.4151534126487165e-05,
2177
+ "loss": 0.0946,
2178
+ "step": 2790
2179
+ },
2180
+ {
2181
+ "epoch": 0.7579859231185706,
2182
+ "grad_norm": 1.71875,
2183
+ "learning_rate": 1.3994990607388856e-05,
2184
+ "loss": 0.0985,
2185
+ "step": 2800
2186
+ },
2187
+ {
2188
+ "epoch": 0.7579859231185706,
2189
+ "eval_loss": 0.13054493069648743,
2190
+ "eval_runtime": 102.656,
2191
+ "eval_samples_per_second": 4.988,
2192
+ "eval_steps_per_second": 0.156,
2193
+ "step": 2800
2194
+ },
2195
+ {
2196
+ "epoch": 0.760693015701137,
2197
+ "grad_norm": 1.1875,
2198
+ "learning_rate": 1.3838447088290547e-05,
2199
+ "loss": 0.1329,
2200
+ "step": 2810
2201
+ },
2202
+ {
2203
+ "epoch": 0.7634001082837033,
2204
+ "grad_norm": 2.21875,
2205
+ "learning_rate": 1.3681903569192236e-05,
2206
+ "loss": 0.1072,
2207
+ "step": 2820
2208
+ },
2209
+ {
2210
+ "epoch": 0.7661072008662696,
2211
+ "grad_norm": 1.390625,
2212
+ "learning_rate": 1.3525360050093926e-05,
2213
+ "loss": 0.0953,
2214
+ "step": 2830
2215
+ },
2216
+ {
2217
+ "epoch": 0.7688142934488359,
2218
+ "grad_norm": 1.046875,
2219
+ "learning_rate": 1.3368816530995617e-05,
2220
+ "loss": 0.0912,
2221
+ "step": 2840
2222
+ },
2223
+ {
2224
+ "epoch": 0.7715213860314023,
2225
+ "grad_norm": 0.984375,
2226
+ "learning_rate": 1.3212273011897308e-05,
2227
+ "loss": 0.1132,
2228
+ "step": 2850
2229
+ },
2230
+ {
2231
+ "epoch": 0.7742284786139686,
2232
+ "grad_norm": 0.5234375,
2233
+ "learning_rate": 1.3055729492798999e-05,
2234
+ "loss": 0.1059,
2235
+ "step": 2860
2236
+ },
2237
+ {
2238
+ "epoch": 0.7769355711965349,
2239
+ "grad_norm": 0.99609375,
2240
+ "learning_rate": 1.289918597370069e-05,
2241
+ "loss": 0.1182,
2242
+ "step": 2870
2243
+ },
2244
+ {
2245
+ "epoch": 0.7796426637791013,
2246
+ "grad_norm": 0.921875,
2247
+ "learning_rate": 1.274264245460238e-05,
2248
+ "loss": 0.0647,
2249
+ "step": 2880
2250
+ },
2251
+ {
2252
+ "epoch": 0.7823497563616676,
2253
+ "grad_norm": 1.609375,
2254
+ "learning_rate": 1.258609893550407e-05,
2255
+ "loss": 0.1049,
2256
+ "step": 2890
2257
+ },
2258
+ {
2259
+ "epoch": 0.7850568489442339,
2260
+ "grad_norm": 2.53125,
2261
+ "learning_rate": 1.2429555416405761e-05,
2262
+ "loss": 0.1223,
2263
+ "step": 2900
2264
+ },
2265
+ {
2266
+ "epoch": 0.7850568489442339,
2267
+ "eval_loss": 0.13052764534950256,
2268
+ "eval_runtime": 114.8255,
2269
+ "eval_samples_per_second": 4.459,
2270
+ "eval_steps_per_second": 0.139,
2271
+ "step": 2900
2272
+ },
2273
+ {
2274
+ "epoch": 0.7877639415268002,
2275
+ "grad_norm": 1.59375,
2276
+ "learning_rate": 1.2273011897307452e-05,
2277
+ "loss": 0.0768,
2278
+ "step": 2910
2279
+ },
2280
+ {
2281
+ "epoch": 0.7904710341093666,
2282
+ "grad_norm": 1.0,
2283
+ "learning_rate": 1.2116468378209143e-05,
2284
+ "loss": 0.0902,
2285
+ "step": 2920
2286
+ },
2287
+ {
2288
+ "epoch": 0.7931781266919329,
2289
+ "grad_norm": 1.0546875,
2290
+ "learning_rate": 1.1959924859110834e-05,
2291
+ "loss": 0.1059,
2292
+ "step": 2930
2293
+ },
2294
+ {
2295
+ "epoch": 0.7958852192744992,
2296
+ "grad_norm": 0.66796875,
2297
+ "learning_rate": 1.1803381340012524e-05,
2298
+ "loss": 0.0919,
2299
+ "step": 2940
2300
+ },
2301
+ {
2302
+ "epoch": 0.7985923118570655,
2303
+ "grad_norm": 1.859375,
2304
+ "learning_rate": 1.1646837820914215e-05,
2305
+ "loss": 0.0779,
2306
+ "step": 2950
2307
+ },
2308
+ {
2309
+ "epoch": 0.8012994044396319,
2310
+ "grad_norm": 1.1328125,
2311
+ "learning_rate": 1.1490294301815906e-05,
2312
+ "loss": 0.1019,
2313
+ "step": 2960
2314
+ },
2315
+ {
2316
+ "epoch": 0.8040064970221982,
2317
+ "grad_norm": 0.8671875,
2318
+ "learning_rate": 1.1333750782717596e-05,
2319
+ "loss": 0.0909,
2320
+ "step": 2970
2321
+ },
2322
+ {
2323
+ "epoch": 0.8067135896047645,
2324
+ "grad_norm": 0.435546875,
2325
+ "learning_rate": 1.1177207263619287e-05,
2326
+ "loss": 0.0973,
2327
+ "step": 2980
2328
+ },
2329
+ {
2330
+ "epoch": 0.8094206821873308,
2331
+ "grad_norm": 1.453125,
2332
+ "learning_rate": 1.1020663744520978e-05,
2333
+ "loss": 0.0875,
2334
+ "step": 2990
2335
+ },
2336
+ {
2337
+ "epoch": 0.8121277747698972,
2338
+ "grad_norm": 0.71484375,
2339
+ "learning_rate": 1.0864120225422668e-05,
2340
+ "loss": 0.1053,
2341
+ "step": 3000
2342
+ },
2343
+ {
2344
+ "epoch": 0.8121277747698972,
2345
+ "eval_loss": 0.13060268759727478,
2346
+ "eval_runtime": 108.9657,
2347
+ "eval_samples_per_second": 4.699,
2348
+ "eval_steps_per_second": 0.147,
2349
+ "step": 3000
2350
+ },
2351
+ {
2352
+ "epoch": 0.8148348673524635,
2353
+ "grad_norm": 1.140625,
2354
+ "learning_rate": 1.070757670632436e-05,
2355
+ "loss": 0.0808,
2356
+ "step": 3010
2357
+ },
2358
+ {
2359
+ "epoch": 0.8175419599350298,
2360
+ "grad_norm": 2.140625,
2361
+ "learning_rate": 1.0551033187226048e-05,
2362
+ "loss": 0.0766,
2363
+ "step": 3020
2364
+ },
2365
+ {
2366
+ "epoch": 0.8202490525175961,
2367
+ "grad_norm": 1.0546875,
2368
+ "learning_rate": 1.0394489668127739e-05,
2369
+ "loss": 0.073,
2370
+ "step": 3030
2371
+ },
2372
+ {
2373
+ "epoch": 0.8229561451001625,
2374
+ "grad_norm": 1.859375,
2375
+ "learning_rate": 1.023794614902943e-05,
2376
+ "loss": 0.108,
2377
+ "step": 3040
2378
+ },
2379
+ {
2380
+ "epoch": 0.8256632376827288,
2381
+ "grad_norm": 0.69140625,
2382
+ "learning_rate": 1.008140262993112e-05,
2383
+ "loss": 0.0815,
2384
+ "step": 3050
2385
+ },
2386
+ {
2387
+ "epoch": 0.828370330265295,
2388
+ "grad_norm": 0.80859375,
2389
+ "learning_rate": 9.924859110832813e-06,
2390
+ "loss": 0.0965,
2391
+ "step": 3060
2392
+ },
2393
+ {
2394
+ "epoch": 0.8310774228478613,
2395
+ "grad_norm": 1.953125,
2396
+ "learning_rate": 9.768315591734503e-06,
2397
+ "loss": 0.1145,
2398
+ "step": 3070
2399
+ },
2400
+ {
2401
+ "epoch": 0.8337845154304278,
2402
+ "grad_norm": 0.734375,
2403
+ "learning_rate": 9.611772072636194e-06,
2404
+ "loss": 0.1319,
2405
+ "step": 3080
2406
+ },
2407
+ {
2408
+ "epoch": 0.836491608012994,
2409
+ "grad_norm": 1.671875,
2410
+ "learning_rate": 9.455228553537885e-06,
2411
+ "loss": 0.0871,
2412
+ "step": 3090
2413
+ },
2414
+ {
2415
+ "epoch": 0.8391987005955603,
2416
+ "grad_norm": 1.515625,
2417
+ "learning_rate": 9.298685034439576e-06,
2418
+ "loss": 0.1207,
2419
+ "step": 3100
2420
+ },
2421
+ {
2422
+ "epoch": 0.8391987005955603,
2423
+ "eval_loss": 0.1296384036540985,
2424
+ "eval_runtime": 103.7944,
2425
+ "eval_samples_per_second": 4.933,
2426
+ "eval_steps_per_second": 0.154,
2427
+ "step": 3100
2428
+ },
2429
+ {
2430
+ "epoch": 0.8419057931781266,
2431
+ "grad_norm": 2.203125,
2432
+ "learning_rate": 9.142141515341266e-06,
2433
+ "loss": 0.1177,
2434
+ "step": 3110
2435
+ },
2436
+ {
2437
+ "epoch": 0.844612885760693,
2438
+ "grad_norm": 1.1171875,
2439
+ "learning_rate": 8.985597996242955e-06,
2440
+ "loss": 0.1111,
2441
+ "step": 3120
2442
+ },
2443
+ {
2444
+ "epoch": 0.8473199783432593,
2445
+ "grad_norm": 0.384765625,
2446
+ "learning_rate": 8.829054477144646e-06,
2447
+ "loss": 0.0774,
2448
+ "step": 3130
2449
+ },
2450
+ {
2451
+ "epoch": 0.8500270709258256,
2452
+ "grad_norm": 1.3984375,
2453
+ "learning_rate": 8.672510958046337e-06,
2454
+ "loss": 0.0864,
2455
+ "step": 3140
2456
+ },
2457
+ {
2458
+ "epoch": 0.852734163508392,
2459
+ "grad_norm": 0.42578125,
2460
+ "learning_rate": 8.515967438948027e-06,
2461
+ "loss": 0.0914,
2462
+ "step": 3150
2463
+ },
2464
+ {
2465
+ "epoch": 0.8554412560909583,
2466
+ "grad_norm": 0.453125,
2467
+ "learning_rate": 8.359423919849718e-06,
2468
+ "loss": 0.0947,
2469
+ "step": 3160
2470
+ },
2471
+ {
2472
+ "epoch": 0.8581483486735246,
2473
+ "grad_norm": 1.265625,
2474
+ "learning_rate": 8.202880400751409e-06,
2475
+ "loss": 0.0812,
2476
+ "step": 3170
2477
+ },
2478
+ {
2479
+ "epoch": 0.8608554412560909,
2480
+ "grad_norm": 1.296875,
2481
+ "learning_rate": 8.0463368816531e-06,
2482
+ "loss": 0.0924,
2483
+ "step": 3180
2484
+ },
2485
+ {
2486
+ "epoch": 0.8635625338386573,
2487
+ "grad_norm": 0.69921875,
2488
+ "learning_rate": 7.889793362554792e-06,
2489
+ "loss": 0.1047,
2490
+ "step": 3190
2491
+ },
2492
+ {
2493
+ "epoch": 0.8662696264212236,
2494
+ "grad_norm": 2.0,
2495
+ "learning_rate": 7.733249843456483e-06,
2496
+ "loss": 0.1071,
2497
+ "step": 3200
2498
+ },
2499
+ {
2500
+ "epoch": 0.8662696264212236,
2501
+ "eval_loss": 0.128614142537117,
2502
+ "eval_runtime": 105.0877,
2503
+ "eval_samples_per_second": 4.872,
2504
+ "eval_steps_per_second": 0.152,
2505
+ "step": 3200
2506
+ },
2507
+ {
2508
+ "epoch": 0.8689767190037899,
2509
+ "grad_norm": 1.8984375,
2510
+ "learning_rate": 7.576706324358172e-06,
2511
+ "loss": 0.0834,
2512
+ "step": 3210
2513
+ },
2514
+ {
2515
+ "epoch": 0.8716838115863562,
2516
+ "grad_norm": 0.94921875,
2517
+ "learning_rate": 7.420162805259863e-06,
2518
+ "loss": 0.1098,
2519
+ "step": 3220
2520
+ },
2521
+ {
2522
+ "epoch": 0.8743909041689226,
2523
+ "grad_norm": 1.2890625,
2524
+ "learning_rate": 7.263619286161554e-06,
2525
+ "loss": 0.1152,
2526
+ "step": 3230
2527
+ },
2528
+ {
2529
+ "epoch": 0.8770979967514889,
2530
+ "grad_norm": 0.63671875,
2531
+ "learning_rate": 7.107075767063244e-06,
2532
+ "loss": 0.1093,
2533
+ "step": 3240
2534
+ },
2535
+ {
2536
+ "epoch": 0.8798050893340552,
2537
+ "grad_norm": 0.93359375,
2538
+ "learning_rate": 6.950532247964934e-06,
2539
+ "loss": 0.1116,
2540
+ "step": 3250
2541
+ },
2542
+ {
2543
+ "epoch": 0.8825121819166215,
2544
+ "grad_norm": 1.421875,
2545
+ "learning_rate": 6.793988728866625e-06,
2546
+ "loss": 0.0824,
2547
+ "step": 3260
2548
+ },
2549
+ {
2550
+ "epoch": 0.8852192744991879,
2551
+ "grad_norm": 1.6953125,
2552
+ "learning_rate": 6.637445209768316e-06,
2553
+ "loss": 0.1241,
2554
+ "step": 3270
2555
+ },
2556
+ {
2557
+ "epoch": 0.8879263670817542,
2558
+ "grad_norm": 1.0859375,
2559
+ "learning_rate": 6.4809016906700065e-06,
2560
+ "loss": 0.0967,
2561
+ "step": 3280
2562
+ },
2563
+ {
2564
+ "epoch": 0.8906334596643205,
2565
+ "grad_norm": 1.125,
2566
+ "learning_rate": 6.324358171571697e-06,
2567
+ "loss": 0.0893,
2568
+ "step": 3290
2569
+ },
2570
+ {
2571
+ "epoch": 0.8933405522468868,
2572
+ "grad_norm": 1.1875,
2573
+ "learning_rate": 6.167814652473388e-06,
2574
+ "loss": 0.1081,
2575
+ "step": 3300
2576
+ },
2577
+ {
2578
+ "epoch": 0.8933405522468868,
2579
+ "eval_loss": 0.12801626324653625,
2580
+ "eval_runtime": 103.0737,
2581
+ "eval_samples_per_second": 4.967,
2582
+ "eval_steps_per_second": 0.155,
2583
+ "step": 3300
2584
+ },
2585
+ {
2586
+ "epoch": 0.8960476448294532,
2587
+ "grad_norm": 1.3984375,
2588
+ "learning_rate": 6.011271133375079e-06,
2589
+ "loss": 0.1199,
2590
+ "step": 3310
2591
+ },
2592
+ {
2593
+ "epoch": 0.8987547374120195,
2594
+ "grad_norm": 1.640625,
2595
+ "learning_rate": 5.854727614276769e-06,
2596
+ "loss": 0.0997,
2597
+ "step": 3320
2598
+ },
2599
+ {
2600
+ "epoch": 0.9014618299945858,
2601
+ "grad_norm": 0.921875,
2602
+ "learning_rate": 5.69818409517846e-06,
2603
+ "loss": 0.1118,
2604
+ "step": 3330
2605
+ },
2606
+ {
2607
+ "epoch": 0.9041689225771521,
2608
+ "grad_norm": 1.3828125,
2609
+ "learning_rate": 5.54164057608015e-06,
2610
+ "loss": 0.083,
2611
+ "step": 3340
2612
+ },
2613
+ {
2614
+ "epoch": 0.9068760151597185,
2615
+ "grad_norm": 1.609375,
2616
+ "learning_rate": 5.3850970569818414e-06,
2617
+ "loss": 0.0765,
2618
+ "step": 3350
2619
+ },
2620
+ {
2621
+ "epoch": 0.9095831077422848,
2622
+ "grad_norm": 1.0546875,
2623
+ "learning_rate": 5.228553537883532e-06,
2624
+ "loss": 0.0925,
2625
+ "step": 3360
2626
+ },
2627
+ {
2628
+ "epoch": 0.9122902003248511,
2629
+ "grad_norm": 1.203125,
2630
+ "learning_rate": 5.072010018785223e-06,
2631
+ "loss": 0.0941,
2632
+ "step": 3370
2633
+ },
2634
+ {
2635
+ "epoch": 0.9149972929074174,
2636
+ "grad_norm": 1.1796875,
2637
+ "learning_rate": 4.9154664996869136e-06,
2638
+ "loss": 0.1193,
2639
+ "step": 3380
2640
+ },
2641
+ {
2642
+ "epoch": 0.9177043854899838,
2643
+ "grad_norm": 0.333984375,
2644
+ "learning_rate": 4.758922980588603e-06,
2645
+ "loss": 0.1124,
2646
+ "step": 3390
2647
+ },
2648
+ {
2649
+ "epoch": 0.9204114780725501,
2650
+ "grad_norm": 1.6796875,
2651
+ "learning_rate": 4.602379461490294e-06,
2652
+ "loss": 0.0924,
2653
+ "step": 3400
2654
+ },
2655
+ {
2656
+ "epoch": 0.9204114780725501,
2657
+ "eval_loss": 0.12878485023975372,
2658
+ "eval_runtime": 123.0744,
2659
+ "eval_samples_per_second": 4.16,
2660
+ "eval_steps_per_second": 0.13,
2661
+ "step": 3400
2662
+ },
2663
+ {
2664
+ "epoch": 0.9231185706551164,
2665
+ "grad_norm": 0.86328125,
2666
+ "learning_rate": 4.445835942391985e-06,
2667
+ "loss": 0.0938,
2668
+ "step": 3410
2669
+ },
2670
+ {
2671
+ "epoch": 0.9258256632376828,
2672
+ "grad_norm": 2.328125,
2673
+ "learning_rate": 4.289292423293676e-06,
2674
+ "loss": 0.0839,
2675
+ "step": 3420
2676
+ },
2677
+ {
2678
+ "epoch": 0.9285327558202491,
2679
+ "grad_norm": 1.7578125,
2680
+ "learning_rate": 4.132748904195367e-06,
2681
+ "loss": 0.1368,
2682
+ "step": 3430
2683
+ },
2684
+ {
2685
+ "epoch": 0.9312398484028154,
2686
+ "grad_norm": 2.515625,
2687
+ "learning_rate": 3.976205385097057e-06,
2688
+ "loss": 0.1154,
2689
+ "step": 3440
2690
+ },
2691
+ {
2692
+ "epoch": 0.9339469409853817,
2693
+ "grad_norm": 0.205078125,
2694
+ "learning_rate": 3.819661865998748e-06,
2695
+ "loss": 0.095,
2696
+ "step": 3450
2697
+ },
2698
+ {
2699
+ "epoch": 0.9366540335679481,
2700
+ "grad_norm": 0.87890625,
2701
+ "learning_rate": 3.6631183469004384e-06,
2702
+ "loss": 0.0872,
2703
+ "step": 3460
2704
+ },
2705
+ {
2706
+ "epoch": 0.9393611261505144,
2707
+ "grad_norm": 2.515625,
2708
+ "learning_rate": 3.506574827802129e-06,
2709
+ "loss": 0.092,
2710
+ "step": 3470
2711
+ },
2712
+ {
2713
+ "epoch": 0.9420682187330807,
2714
+ "grad_norm": 1.125,
2715
+ "learning_rate": 3.35003130870382e-06,
2716
+ "loss": 0.0944,
2717
+ "step": 3480
2718
+ },
2719
+ {
2720
+ "epoch": 0.944775311315647,
2721
+ "grad_norm": 1.578125,
2722
+ "learning_rate": 3.193487789605511e-06,
2723
+ "loss": 0.1366,
2724
+ "step": 3490
2725
+ },
2726
+ {
2727
+ "epoch": 0.9474824038982134,
2728
+ "grad_norm": 0.5859375,
2729
+ "learning_rate": 3.036944270507201e-06,
2730
+ "loss": 0.0948,
2731
+ "step": 3500
2732
+ },
2733
+ {
2734
+ "epoch": 0.9474824038982134,
2735
+ "eval_loss": 0.1291881799697876,
2736
+ "eval_runtime": 104.6704,
2737
+ "eval_samples_per_second": 4.892,
2738
+ "eval_steps_per_second": 0.153,
2739
+ "step": 3500
2740
+ },
2741
+ {
2742
+ "epoch": 0.9501894964807797,
2743
+ "grad_norm": 1.375,
2744
+ "learning_rate": 2.880400751408892e-06,
2745
+ "loss": 0.1181,
2746
+ "step": 3510
2747
+ },
2748
+ {
2749
+ "epoch": 0.952896589063346,
2750
+ "grad_norm": 0.84375,
2751
+ "learning_rate": 2.7238572323105826e-06,
2752
+ "loss": 0.0819,
2753
+ "step": 3520
2754
+ },
2755
+ {
2756
+ "epoch": 0.9556036816459123,
2757
+ "grad_norm": 0.9765625,
2758
+ "learning_rate": 2.5673137132122733e-06,
2759
+ "loss": 0.1134,
2760
+ "step": 3530
2761
+ },
2762
+ {
2763
+ "epoch": 0.9583107742284787,
2764
+ "grad_norm": 0.5625,
2765
+ "learning_rate": 2.410770194113964e-06,
2766
+ "loss": 0.1183,
2767
+ "step": 3540
2768
+ },
2769
+ {
2770
+ "epoch": 0.961017866811045,
2771
+ "grad_norm": 0.84765625,
2772
+ "learning_rate": 2.2542266750156543e-06,
2773
+ "loss": 0.1021,
2774
+ "step": 3550
2775
+ },
2776
+ {
2777
+ "epoch": 0.9637249593936112,
2778
+ "grad_norm": 1.0390625,
2779
+ "learning_rate": 2.0976831559173454e-06,
2780
+ "loss": 0.0871,
2781
+ "step": 3560
2782
+ },
2783
+ {
2784
+ "epoch": 0.9664320519761775,
2785
+ "grad_norm": 1.1171875,
2786
+ "learning_rate": 1.9411396368190357e-06,
2787
+ "loss": 0.1036,
2788
+ "step": 3570
2789
+ },
2790
+ {
2791
+ "epoch": 0.969139144558744,
2792
+ "grad_norm": 0.546875,
2793
+ "learning_rate": 1.7845961177207264e-06,
2794
+ "loss": 0.0874,
2795
+ "step": 3580
2796
+ },
2797
+ {
2798
+ "epoch": 0.9718462371413102,
2799
+ "grad_norm": 0.28515625,
2800
+ "learning_rate": 1.6280525986224169e-06,
2801
+ "loss": 0.085,
2802
+ "step": 3590
2803
+ },
2804
+ {
2805
+ "epoch": 0.9745533297238765,
2806
+ "grad_norm": 0.90234375,
2807
+ "learning_rate": 1.4715090795241078e-06,
2808
+ "loss": 0.1162,
2809
+ "step": 3600
2810
+ },
2811
+ {
2812
+ "epoch": 0.9745533297238765,
2813
+ "eval_loss": 0.12912487983703613,
2814
+ "eval_runtime": 106.4924,
2815
+ "eval_samples_per_second": 4.808,
2816
+ "eval_steps_per_second": 0.15,
2817
+ "step": 3600
2818
+ },
2819
+ {
2820
+ "epoch": 0.9772604223064428,
2821
+ "grad_norm": 0.74609375,
2822
+ "learning_rate": 1.3149655604257985e-06,
2823
+ "loss": 0.1066,
2824
+ "step": 3610
2825
+ },
2826
+ {
2827
+ "epoch": 0.9799675148890092,
2828
+ "grad_norm": 1.4609375,
2829
+ "learning_rate": 1.1584220413274892e-06,
2830
+ "loss": 0.1533,
2831
+ "step": 3620
2832
+ },
2833
+ {
2834
+ "epoch": 0.9826746074715755,
2835
+ "grad_norm": 0.68359375,
2836
+ "learning_rate": 1.0018785222291797e-06,
2837
+ "loss": 0.0717,
2838
+ "step": 3630
2839
+ },
2840
+ {
2841
+ "epoch": 0.9853817000541418,
2842
+ "grad_norm": 0.94921875,
2843
+ "learning_rate": 8.453350031308704e-07,
2844
+ "loss": 0.0667,
2845
+ "step": 3640
2846
+ },
2847
+ {
2848
+ "epoch": 0.9880887926367081,
2849
+ "grad_norm": 0.427734375,
2850
+ "learning_rate": 6.887914840325611e-07,
2851
+ "loss": 0.1175,
2852
+ "step": 3650
2853
+ },
2854
+ {
2855
+ "epoch": 0.9907958852192745,
2856
+ "grad_norm": 1.046875,
2857
+ "learning_rate": 5.322479649342517e-07,
2858
+ "loss": 0.0871,
2859
+ "step": 3660
2860
+ },
2861
+ {
2862
+ "epoch": 0.9935029778018408,
2863
+ "grad_norm": 1.8515625,
2864
+ "learning_rate": 3.757044458359424e-07,
2865
+ "loss": 0.122,
2866
+ "step": 3670
2867
+ },
2868
+ {
2869
+ "epoch": 0.9962100703844071,
2870
+ "grad_norm": 0.77734375,
2871
+ "learning_rate": 2.1916092673763307e-07,
2872
+ "loss": 0.0922,
2873
+ "step": 3680
2874
+ },
2875
+ {
2876
+ "epoch": 0.9989171629669734,
2877
+ "grad_norm": 3.140625,
2878
+ "learning_rate": 6.261740763932373e-08,
2879
+ "loss": 0.1331,
2880
+ "step": 3690
2881
+ }
2882
+ ],
2883
+ "logging_steps": 10,
2884
+ "max_steps": 3694,
2885
+ "num_input_tokens_seen": 0,
2886
+ "num_train_epochs": 1,
2887
+ "save_steps": 500,
2888
+ "stateful_callbacks": {
2889
+ "TrainerControl": {
2890
+ "args": {
2891
+ "should_epoch_stop": false,
2892
+ "should_evaluate": false,
2893
+ "should_log": false,
2894
+ "should_save": true,
2895
+ "should_training_stop": true
2896
+ },
2897
+ "attributes": {}
2898
+ }
2899
+ },
2900
+ "total_flos": 1.0990508415397056e+18,
2901
+ "train_batch_size": 32,
2902
+ "trial_name": null,
2903
+ "trial_params": null
2904
+ }
checkpoint-3694/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:685ee2596072507a9c0ce1ae768b36fc3337189cbe14f6c946fc5fb838a3c04b
3
+ size 5176
git_hash.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ a5b4a24649e45a6570b33cc01acb57807e4dfa5b
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
preprocessor_config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_convert_rgb": true,
3
+ "do_normalize": true,
4
+ "do_rescale": true,
5
+ "do_resize": true,
6
+ "image_mean": [
7
+ 0.48145466,
8
+ 0.4578275,
9
+ 0.40821073
10
+ ],
11
+ "image_processor_type": "Qwen2VLImageProcessor",
12
+ "image_std": [
13
+ 0.26862954,
14
+ 0.26130258,
15
+ 0.27577711
16
+ ],
17
+ "max_pixels": 12845056,
18
+ "merge_size": 2,
19
+ "min_pixels": 3136,
20
+ "patch_size": 14,
21
+ "processor_class": "ColQwen2Processor",
22
+ "resample": 3,
23
+ "rescale_factor": 0.00392156862745098,
24
+ "size": {
25
+ "max_pixels": 12845056,
26
+ "min_pixels": 3136
27
+ },
28
+ "temporal_patch_size": 2
29
+ }
results.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"validation_set": {"ndcg_at_1": 0.81641, "ndcg_at_3": 0.87151, "ndcg_at_5": 0.87924, "ndcg_at_10": 0.88352, "ndcg_at_20": 0.88599, "ndcg_at_100": 0.89228, "ndcg_at_1000": 0.8945, "map_at_1": 0.81641, "map_at_3": 0.85872, "map_at_5": 0.86283, "map_at_10": 0.86451, "map_at_20": 0.86519, "map_at_100": 0.86613, "map_at_1000": 0.86625, "recall_at_1": 0.81641, "recall_at_3": 0.9082, "recall_at_5": 0.92773, "recall_at_10": 0.94141, "recall_at_20": 0.95117, "recall_at_100": 0.98438, "recall_at_1000": 1.0, "precision_at_1": 0.81641, "precision_at_3": 0.30273, "precision_at_5": 0.18555, "precision_at_10": 0.09414, "precision_at_20": 0.04756, "precision_at_100": 0.00984, "precision_at_1000": 0.001, "mrr_at_1": 0.8203125, "mrr_at_3": 0.8603515624999998, "mrr_at_5": 0.8628906249999998, "mrr_at_10": 0.8655544704861109, "mrr_at_20": 0.8665781348642675, "mrr_at_100": 0.8675737791279903, "mrr_at_1000": 0.8676711773156423, "naucs_at_1_max": 0.34667870891415986, "naucs_at_1_std": -0.20661616231925453, "naucs_at_1_diff1": 0.9396571084820124, "naucs_at_3_max": 0.5166729885733626, "naucs_at_3_std": 0.05950323676330907, "naucs_at_3_diff1": 0.9342216236825106, "naucs_at_5_max": 0.6329364969562152, "naucs_at_5_std": 0.2619697444169589, "naucs_at_5_diff1": 0.9376321560083725, "naucs_at_10_max": 0.660722672615821, "naucs_at_10_std": 0.437248626444334, "naucs_at_10_diff1": 0.9546510563783107, "naucs_at_20_max": 0.6658122614005673, "naucs_at_20_std": 0.4847758170885219, "naucs_at_20_diff1": 0.9508077573821819, "naucs_at_100_max": 0.6851951655323608, "naucs_at_100_std": 0.8932409444836575, "naucs_at_100_diff1": 0.9510016587980088, "naucs_at_1000_max": 1.0, "naucs_at_1000_std": 1.0, "naucs_at_1000_diff1": 1.0}, "syntheticDocQA_energy": {"ndcg_at_1": 0.92, "ndcg_at_3": 0.94762, "ndcg_at_5": 0.94762, "ndcg_at_10": 0.94762, "ndcg_at_20": 0.95512, "ndcg_at_100": 0.95512, "ndcg_at_1000": 0.95512, "map_at_1": 0.92, "map_at_3": 0.94, "map_at_5": 0.94, "map_at_10": 0.94, "map_at_20": 0.94201, "map_at_100": 0.94201, "map_at_1000": 0.94201, "recall_at_1": 0.92, "recall_at_3": 0.97, "recall_at_5": 0.97, "recall_at_10": 0.97, "recall_at_20": 1.0, "recall_at_100": 1.0, "recall_at_1000": 1.0, "precision_at_1": 0.92, "precision_at_3": 0.32333, "precision_at_5": 0.194, "precision_at_10": 0.097, "precision_at_20": 0.05, "precision_at_100": 0.01, "precision_at_1000": 0.001, "mrr_at_1": 0.92, "mrr_at_3": 0.94, "mrr_at_5": 0.94, "mrr_at_10": 0.94, "mrr_at_20": 0.9421501831501832, "mrr_at_100": 0.9421501831501832, "mrr_at_1000": 0.9421501831501832, "naucs_at_1_max": 0.5697362278244628, "naucs_at_1_std": -0.5218253968253984, "naucs_at_1_diff1": 0.9489379084967322, "naucs_at_3_max": 0.7642390289449176, "naucs_at_3_std": -0.8249299719887898, "naucs_at_3_diff1": 1.0, "naucs_at_5_max": 0.7642390289449118, "naucs_at_5_std": -0.8249299719887969, "naucs_at_5_diff1": 1.0, "naucs_at_10_max": 0.7642390289449118, "naucs_at_10_std": -0.8249299719887969, "naucs_at_10_diff1": 1.0, "naucs_at_20_max": 1.0, "naucs_at_20_std": 1.0, "naucs_at_20_diff1": 1.0, "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": 0.93, "ndcg_at_3": 0.97286, "ndcg_at_5": 0.97286, "ndcg_at_10": 0.97286, "ndcg_at_20": 0.97286, "ndcg_at_100": 0.97286, "ndcg_at_1000": 0.97286, "map_at_1": 0.93, "map_at_3": 0.96333, "map_at_5": 0.96333, "map_at_10": 0.96333, "map_at_20": 0.96333, "map_at_100": 0.96333, "map_at_1000": 0.96333, "recall_at_1": 0.93, "recall_at_3": 1.0, "recall_at_5": 1.0, "recall_at_10": 1.0, "recall_at_20": 1.0, "recall_at_100": 1.0, "recall_at_1000": 1.0, "precision_at_1": 0.93, "precision_at_3": 0.33333, "precision_at_5": 0.2, "precision_at_10": 0.1, "precision_at_20": 0.05, "precision_at_100": 0.01, "precision_at_1000": 0.001, "mrr_at_1": 0.93, "mrr_at_3": 0.9633333333333334, "mrr_at_5": 0.9633333333333334, "mrr_at_10": 0.9633333333333334, "mrr_at_20": 0.9633333333333334, "mrr_at_100": 0.9633333333333334, "mrr_at_1000": 0.9633333333333334, "naucs_at_1_max": 0.7465652927837806, "naucs_at_1_std": 0.10897692410297374, "naucs_at_1_diff1": 0.9626517273576113, "naucs_at_3_max": 1.0, "naucs_at_3_std": 1.0, "naucs_at_3_diff1": 1.0, "naucs_at_5_max": 1.0, "naucs_at_5_std": 1.0, "naucs_at_5_diff1": 1.0, "naucs_at_10_max": 1.0, "naucs_at_10_std": 1.0, "naucs_at_10_diff1": 1.0, "naucs_at_20_max": 1.0, "naucs_at_20_std": 1.0, "naucs_at_20_diff1": 1.0, "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": 0.97, "ndcg_at_3": 0.98762, "ndcg_at_5": 0.98762, "ndcg_at_10": 0.98762, "ndcg_at_20": 0.98762, "ndcg_at_100": 0.98762, "ndcg_at_1000": 0.98762, "map_at_1": 0.97, "map_at_3": 0.98333, "map_at_5": 0.98333, "map_at_10": 0.98333, "map_at_20": 0.98333, "map_at_100": 0.98333, "map_at_1000": 0.98333, "recall_at_1": 0.97, "recall_at_3": 1.0, "recall_at_5": 1.0, "recall_at_10": 1.0, "recall_at_20": 1.0, "recall_at_100": 1.0, "recall_at_1000": 1.0, "precision_at_1": 0.97, "precision_at_3": 0.33333, "precision_at_5": 0.2, "precision_at_10": 0.1, "precision_at_20": 0.05, "precision_at_100": 0.01, "precision_at_1000": 0.001, "mrr_at_1": 0.97, "mrr_at_3": 0.9833333333333333, "mrr_at_5": 0.9833333333333333, "mrr_at_10": 0.9833333333333333, "mrr_at_20": 0.9833333333333333, "mrr_at_100": 0.9833333333333333, "mrr_at_1000": 0.9833333333333333, "naucs_at_1_max": 0.8638344226579567, "naucs_at_1_std": -0.28197945845004707, "naucs_at_1_diff1": 0.9564270152505465, "naucs_at_3_max": 1.0, "naucs_at_3_std": 1.0, "naucs_at_3_diff1": 1.0, "naucs_at_5_max": 1.0, "naucs_at_5_std": 1.0, "naucs_at_5_diff1": 1.0, "naucs_at_10_max": 1.0, "naucs_at_10_std": 1.0, "naucs_at_10_diff1": 1.0, "naucs_at_20_max": 1.0, "naucs_at_20_std": 1.0, "naucs_at_20_diff1": 1.0, "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": 0.85, "ndcg_at_3": 0.92809, "ndcg_at_5": 0.9324, "ndcg_at_10": 0.9324, "ndcg_at_20": 0.93503, "ndcg_at_100": 0.93503, "ndcg_at_1000": 0.93503, "map_at_1": 0.85, "map_at_3": 0.91, "map_at_5": 0.9125, "map_at_10": 0.9125, "map_at_20": 0.91327, "map_at_100": 0.91327, "map_at_1000": 0.91327, "recall_at_1": 0.85, "recall_at_3": 0.98, "recall_at_5": 0.99, "recall_at_10": 0.99, "recall_at_20": 1.0, "recall_at_100": 1.0, "recall_at_1000": 1.0, "precision_at_1": 0.85, "precision_at_3": 0.32667, "precision_at_5": 0.198, "precision_at_10": 0.099, "precision_at_20": 0.05, "precision_at_100": 0.01, "precision_at_1000": 0.001, "mrr_at_1": 0.87, "mrr_at_3": 0.9266666666666667, "mrr_at_5": 0.9266666666666667, "mrr_at_10": 0.9266666666666667, "mrr_at_20": 0.9274358974358975, "mrr_at_100": 0.9274358974358975, "mrr_at_1000": 0.9274358974358975, "naucs_at_1_max": 0.47617962902700983, "naucs_at_1_std": 0.33045883501464407, "naucs_at_1_diff1": 0.8781646599414249, "naucs_at_3_max": 1.0, "naucs_at_3_std": 0.6381886087768379, "naucs_at_3_diff1": 1.0, "naucs_at_5_max": 1.0, "naucs_at_5_std": 0.7222222222222276, "naucs_at_5_diff1": 1.0, "naucs_at_10_max": 1.0, "naucs_at_10_std": 0.7222222222222276, "naucs_at_10_diff1": 1.0, "naucs_at_20_max": 1.0, "naucs_at_20_std": 1.0, "naucs_at_20_diff1": 1.0, "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}, "infovqa_subsampled": {"ndcg_at_1": 0.86, "ndcg_at_3": 0.88671, "ndcg_at_5": 0.89825, "ndcg_at_10": 0.90151, "ndcg_at_20": 0.90849, "ndcg_at_100": 0.91162, "ndcg_at_1000": 0.91339, "map_at_1": 0.86, "map_at_3": 0.88067, "map_at_5": 0.88707, "map_at_10": 0.88843, "map_at_20": 0.89056, "map_at_100": 0.89092, "map_at_1000": 0.89099, "recall_at_1": 0.86, "recall_at_3": 0.904, "recall_at_5": 0.932, "recall_at_10": 0.942, "recall_at_20": 0.968, "recall_at_100": 0.986, "recall_at_1000": 1.0, "precision_at_1": 0.86, "precision_at_3": 0.30133, "precision_at_5": 0.1864, "precision_at_10": 0.0942, "precision_at_20": 0.0484, "precision_at_100": 0.00986, "precision_at_1000": 0.001, "mrr_at_1": 0.858, "mrr_at_3": 0.8799999999999999, "mrr_at_5": 0.8859999999999999, "mrr_at_10": 0.8877992063492062, "mrr_at_20": 0.8897055999555997, "mrr_at_100": 0.8900658496714725, "mrr_at_1000": 0.8901272220832652, "naucs_at_1_max": 0.6271096023278367, "naucs_at_1_std": -0.043875571567133, "naucs_at_1_diff1": 0.9370583344880139, "naucs_at_3_max": 0.6881127450980395, "naucs_at_3_std": -0.021095938375348355, "naucs_at_3_diff1": 0.9083313881108002, "naucs_at_5_max": 0.8094002306805071, "naucs_at_5_std": 0.05795847750865092, "naucs_at_5_diff1": 0.9197424067666282, "naucs_at_10_max": 0.877426832802087, "naucs_at_10_std": 0.2035963810811664, "naucs_at_10_diff1": 0.9250619788145158, "naucs_at_20_max": 0.9557948179271724, "naucs_at_20_std": 0.38013538748832426, "naucs_at_20_diff1": 0.9325980392156829, "naucs_at_100_max": 0.9626517273576126, "naucs_at_100_std": 0.448312658396679, "naucs_at_100_diff1": 0.981325863678799, "naucs_at_1000_max": 1.0, "naucs_at_1000_std": 1.0, "naucs_at_1000_diff1": 1.0}, "docvqa_subsampled": {"ndcg_at_1": 0.47, "ndcg_at_3": 0.53112, "ndcg_at_5": 0.56176, "ndcg_at_10": 0.58036, "ndcg_at_20": 0.59124, "ndcg_at_100": 0.61372, "ndcg_at_1000": 0.62957, "map_at_1": 0.47, "map_at_3": 0.51567, "map_at_5": 0.53277, "map_at_10": 0.54035, "map_at_20": 0.54347, "map_at_100": 0.54625, "map_at_1000": 0.54684, "recall_at_1": 0.47, "recall_at_3": 0.576, "recall_at_5": 0.65, "recall_at_10": 0.708, "recall_at_20": 0.75, "recall_at_100": 0.876, "recall_at_1000": 1.0, "precision_at_1": 0.47, "precision_at_3": 0.192, "precision_at_5": 0.13, "precision_at_10": 0.0708, "precision_at_20": 0.0375, "precision_at_100": 0.00876, "precision_at_1000": 0.001, "mrr_at_1": 0.47, "mrr_at_3": 0.5140000000000002, "mrr_at_5": 0.5322999999999999, "mrr_at_10": 0.5393452380952379, "mrr_at_20": 0.5430581398134026, "mrr_at_100": 0.5455877170766977, "mrr_at_1000": 0.5461605124926291, "naucs_at_1_max": 0.5899734964859769, "naucs_at_1_std": 0.37540315668719265, "naucs_at_1_diff1": 0.8310658104669103, "naucs_at_3_max": 0.5424525576127407, "naucs_at_3_std": 0.425733305596006, "naucs_at_3_diff1": 0.7247359175276799, "naucs_at_5_max": 0.5197754089022696, "naucs_at_5_std": 0.5280852795182682, "naucs_at_5_diff1": 0.6994124827081123, "naucs_at_10_max": 0.4841097869266883, "naucs_at_10_std": 0.5795468783396347, "naucs_at_10_diff1": 0.6604425085310391, "naucs_at_20_max": 0.4785557299843014, "naucs_at_20_std": 0.6286844583987442, "naucs_at_20_diff1": 0.6366300366300363, "naucs_at_100_max": 0.531426717344952, "naucs_at_100_std": 0.8670699298538366, "naucs_at_100_diff1": 0.5190043570965052, "naucs_at_1000_max": 1.0, "naucs_at_1000_std": 1.0, "naucs_at_1000_diff1": 1.0}, "arxivqa_subsampled": {"ndcg_at_1": 0.8, "ndcg_at_3": 0.84812, "ndcg_at_5": 0.86369, "ndcg_at_10": 0.87151, "ndcg_at_20": 0.87714, "ndcg_at_100": 0.88221, "ndcg_at_1000": 0.88366, "map_at_1": 0.8, "map_at_3": 0.837, "map_at_5": 0.8456, "map_at_10": 0.84885, "map_at_20": 0.85044, "map_at_100": 0.85123, "map_at_1000": 0.85132, "recall_at_1": 0.8, "recall_at_3": 0.88, "recall_at_5": 0.918, "recall_at_10": 0.942, "recall_at_20": 0.964, "recall_at_100": 0.99, "recall_at_1000": 1.0, "precision_at_1": 0.8, "precision_at_3": 0.29333, "precision_at_5": 0.1836, "precision_at_10": 0.0942, "precision_at_20": 0.0482, "precision_at_100": 0.0099, "precision_at_1000": 0.001, "mrr_at_1": 0.794, "mrr_at_3": 0.8339999999999999, "mrr_at_5": 0.8417999999999999, "mrr_at_10": 0.8457571428571429, "mrr_at_20": 0.8470214830491145, "mrr_at_100": 0.847837729621968, "mrr_at_1000": 0.847903412604169, "naucs_at_1_max": 0.7301598401598397, "naucs_at_1_std": 0.13073426573426558, "naucs_at_1_diff1": 0.9151098901098901, "naucs_at_3_max": 0.7855154311167684, "naucs_at_3_std": 0.19653197581928136, "naucs_at_3_diff1": 0.8981466751511297, "naucs_at_5_max": 0.8119719432488443, "naucs_at_5_std": 0.1740338411787435, "naucs_at_5_diff1": 0.9070278517911231, "naucs_at_10_max": 0.8026658939437882, "naucs_at_10_std": 0.2013587044013021, "naucs_at_10_diff1": 0.9108149006729143, "naucs_at_20_max": 0.9162257495590788, "naucs_at_20_std": 0.37184873949579655, "naucs_at_20_diff1": 0.9110384894698595, "naucs_at_100_max": 0.947712418300658, "naucs_at_100_std": 0.3913165266106556, "naucs_at_100_diff1": 0.9738562091503188, "naucs_at_1000_max": 1.0, "naucs_at_1000_std": 1.0, "naucs_at_1000_diff1": 1.0}, "tabfquad_subsampled": {"ndcg_at_1": 0.82143, "ndcg_at_3": 0.88125, "ndcg_at_5": 0.88724, "ndcg_at_10": 0.89646, "ndcg_at_20": 0.90181, "ndcg_at_100": 0.90464, "ndcg_at_1000": 0.90464, "map_at_1": 0.82143, "map_at_3": 0.86726, "map_at_5": 0.87065, "map_at_10": 0.87444, "map_at_20": 0.87587, "map_at_100": 0.87633, "map_at_1000": 0.87633, "recall_at_1": 0.82143, "recall_at_3": 0.92143, "recall_at_5": 0.93571, "recall_at_10": 0.96429, "recall_at_20": 0.98571, "recall_at_100": 1.0, "recall_at_1000": 1.0, "precision_at_1": 0.82143, "precision_at_3": 0.30714, "precision_at_5": 0.18714, "precision_at_10": 0.09643, "precision_at_20": 0.04929, "precision_at_100": 0.01, "precision_at_1000": 0.001, "mrr_at_1": 0.8285714285714286, "mrr_at_3": 0.8708333333333336, "mrr_at_5": 0.8742261904761907, "mrr_at_10": 0.8779606009070298, "mrr_at_20": 0.8793680292629874, "mrr_at_100": 0.8798495301637429, "mrr_at_1000": 0.8798495301637429, "naucs_at_1_max": 0.4043094907866057, "naucs_at_1_std": 0.16813948880523036, "naucs_at_1_diff1": 0.8608282147810575, "naucs_at_3_max": 0.5415287327052029, "naucs_at_3_std": 0.32170443935149834, "naucs_at_3_diff1": 0.7899371869960117, "naucs_at_5_max": 0.5449216723726537, "naucs_at_5_std": 0.295959124390499, "naucs_at_5_diff1": 0.768025728810043, "naucs_at_10_max": 0.6852007469654555, "naucs_at_10_std": 0.596965452847805, "naucs_at_10_diff1": 0.7665732959850585, "naucs_at_20_max": 0.8231792717086873, "naucs_at_20_std": 0.7496498599439745, "naucs_at_20_diff1": 0.6626984126984181, "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": 0.62237, "ndcg_at_3": 0.72528, "ndcg_at_5": 0.7518, "ndcg_at_10": 0.7704, "ndcg_at_20": 0.77829, "ndcg_at_100": 0.78521, "ndcg_at_1000": 0.787, "map_at_1": 0.62237, "map_at_3": 0.70014, "map_at_5": 0.71487, "map_at_10": 0.72262, "map_at_20": 0.72486, "map_at_100": 0.72587, "map_at_1000": 0.72594, "recall_at_1": 0.62237, "recall_at_3": 0.79796, "recall_at_5": 0.8623, "recall_at_10": 0.91942, "recall_at_20": 0.95009, "recall_at_100": 0.98677, "recall_at_1000": 1.0, "precision_at_1": 0.62237, "precision_at_3": 0.26599, "precision_at_5": 0.17246, "precision_at_10": 0.09194, "precision_at_20": 0.0475, "precision_at_100": 0.00987, "precision_at_1000": 0.001, "mrr_at_1": 0.6277811184606134, "mrr_at_3": 0.7027460412908401, "mrr_at_5": 0.7170875927039494, "mrr_at_10": 0.7253199896916075, "mrr_at_20": 0.7277180395724787, "mrr_at_100": 0.7287256043001191, "mrr_at_1000": 0.7287991066528416, "naucs_at_1_max": 0.13553936519738788, "naucs_at_1_std": -0.3187519674953281, "naucs_at_1_diff1": 0.7778249427309417, "naucs_at_3_max": 0.2141299663362708, "naucs_at_3_std": -0.2863201435902689, "naucs_at_3_diff1": 0.6746349432069665, "naucs_at_5_max": 0.26199278239549034, "naucs_at_5_std": -0.2097766943253057, "naucs_at_5_diff1": 0.6504413592115734, "naucs_at_10_max": 0.33005867437923625, "naucs_at_10_std": 0.042015716590848955, "naucs_at_10_diff1": 0.5806124077911619, "naucs_at_20_max": 0.38884886089913584, "naucs_at_20_std": 0.25394752368146895, "naucs_at_20_diff1": 0.576841503912515, "naucs_at_100_max": 0.6555237795463111, "naucs_at_100_std": 0.7055850937111812, "naucs_at_100_diff1": 0.6635432021140835, "naucs_at_1000_max": NaN, "naucs_at_1000_std": NaN, "naucs_at_1000_diff1": NaN}, "shift_project": {"ndcg_at_1": 0.74, "ndcg_at_3": 0.84833, "ndcg_at_5": 0.85651, "ndcg_at_10": 0.86656, "ndcg_at_20": 0.86887, "ndcg_at_100": 0.87251, "ndcg_at_1000": 0.87251, "map_at_1": 0.74, "map_at_3": 0.82333, "map_at_5": 0.82783, "map_at_10": 0.83218, "map_at_20": 0.8327, "map_at_100": 0.83316, "map_at_1000": 0.83316, "recall_at_1": 0.74, "recall_at_3": 0.92, "recall_at_5": 0.94, "recall_at_10": 0.97, "recall_at_20": 0.98, "recall_at_100": 1.0, "recall_at_1000": 1.0, "precision_at_1": 0.74, "precision_at_3": 0.30667, "precision_at_5": 0.188, "precision_at_10": 0.097, "precision_at_20": 0.049, "precision_at_100": 0.01, "precision_at_1000": 0.001, "mrr_at_1": 0.76, "mrr_at_3": 0.8333333333333335, "mrr_at_5": 0.8378333333333334, "mrr_at_10": 0.8423571428571431, "mrr_at_20": 0.8428834586466167, "mrr_at_100": 0.84334179197995, "mrr_at_1000": 0.84334179197995, "naucs_at_1_max": -0.18346782368922665, "naucs_at_1_std": -0.312314991281781, "naucs_at_1_diff1": 0.6844410202343775, "naucs_at_3_max": 0.05304621848739506, "naucs_at_3_std": 0.02485994397759261, "naucs_at_3_diff1": 0.6601890756302518, "naucs_at_5_max": 0.30104263927793357, "naucs_at_5_std": 0.15561780267662587, "naucs_at_5_diff1": 0.7639277933395614, "naucs_at_10_max": -0.35434173669467856, "naucs_at_10_std": -0.3963585434173691, "naucs_at_10_diff1": 0.8638344226579531, "naucs_at_20_max": -0.5929038281979383, "naucs_at_20_std": -0.5088702147525547, "naucs_at_20_diff1": 0.9346405228758136, "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}}
special_tokens_map.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|im_start|>",
4
+ "<|im_end|>",
5
+ "<|object_ref_start|>",
6
+ "<|object_ref_end|>",
7
+ "<|box_start|>",
8
+ "<|box_end|>",
9
+ "<|quad_start|>",
10
+ "<|quad_end|>",
11
+ "<|vision_start|>",
12
+ "<|vision_end|>",
13
+ "<|vision_pad|>",
14
+ "<|image_pad|>",
15
+ "<|video_pad|>"
16
+ ],
17
+ "eos_token": {
18
+ "content": "<|im_end|>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ },
24
+ "pad_token": {
25
+ "content": "<|endoftext|>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ }
31
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,144 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "151643": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": false,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "151644": {
13
+ "content": "<|im_start|>",
14
+ "lstrip": false,
15
+ "normalized": false,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "151645": {
21
+ "content": "<|im_end|>",
22
+ "lstrip": false,
23
+ "normalized": false,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ },
28
+ "151646": {
29
+ "content": "<|object_ref_start|>",
30
+ "lstrip": false,
31
+ "normalized": false,
32
+ "rstrip": false,
33
+ "single_word": false,
34
+ "special": true
35
+ },
36
+ "151647": {
37
+ "content": "<|object_ref_end|>",
38
+ "lstrip": false,
39
+ "normalized": false,
40
+ "rstrip": false,
41
+ "single_word": false,
42
+ "special": true
43
+ },
44
+ "151648": {
45
+ "content": "<|box_start|>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false,
50
+ "special": true
51
+ },
52
+ "151649": {
53
+ "content": "<|box_end|>",
54
+ "lstrip": false,
55
+ "normalized": false,
56
+ "rstrip": false,
57
+ "single_word": false,
58
+ "special": true
59
+ },
60
+ "151650": {
61
+ "content": "<|quad_start|>",
62
+ "lstrip": false,
63
+ "normalized": false,
64
+ "rstrip": false,
65
+ "single_word": false,
66
+ "special": true
67
+ },
68
+ "151651": {
69
+ "content": "<|quad_end|>",
70
+ "lstrip": false,
71
+ "normalized": false,
72
+ "rstrip": false,
73
+ "single_word": false,
74
+ "special": true
75
+ },
76
+ "151652": {
77
+ "content": "<|vision_start|>",
78
+ "lstrip": false,
79
+ "normalized": false,
80
+ "rstrip": false,
81
+ "single_word": false,
82
+ "special": true
83
+ },
84
+ "151653": {
85
+ "content": "<|vision_end|>",
86
+ "lstrip": false,
87
+ "normalized": false,
88
+ "rstrip": false,
89
+ "single_word": false,
90
+ "special": true
91
+ },
92
+ "151654": {
93
+ "content": "<|vision_pad|>",
94
+ "lstrip": false,
95
+ "normalized": false,
96
+ "rstrip": false,
97
+ "single_word": false,
98
+ "special": true
99
+ },
100
+ "151655": {
101
+ "content": "<|image_pad|>",
102
+ "lstrip": false,
103
+ "normalized": false,
104
+ "rstrip": false,
105
+ "single_word": false,
106
+ "special": true
107
+ },
108
+ "151656": {
109
+ "content": "<|video_pad|>",
110
+ "lstrip": false,
111
+ "normalized": false,
112
+ "rstrip": false,
113
+ "single_word": false,
114
+ "special": true
115
+ }
116
+ },
117
+ "additional_special_tokens": [
118
+ "<|im_start|>",
119
+ "<|im_end|>",
120
+ "<|object_ref_start|>",
121
+ "<|object_ref_end|>",
122
+ "<|box_start|>",
123
+ "<|box_end|>",
124
+ "<|quad_start|>",
125
+ "<|quad_end|>",
126
+ "<|vision_start|>",
127
+ "<|vision_end|>",
128
+ "<|vision_pad|>",
129
+ "<|image_pad|>",
130
+ "<|video_pad|>"
131
+ ],
132
+ "bos_token": null,
133
+ "chat_template": "{% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n{% endif %}<|im_start|>{{ message['role'] }}\n{% if message['content'] is string %}{{ message['content'] }}<|im_end|>\n{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>\n{% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant\n{% endif %}",
134
+ "clean_up_tokenization_spaces": false,
135
+ "eos_token": "<|im_end|>",
136
+ "errors": "replace",
137
+ "model_max_length": 32768,
138
+ "pad_token": "<|endoftext|>",
139
+ "padding_side": "left",
140
+ "processor_class": "ColQwen2Processor",
141
+ "split_special_tokens": false,
142
+ "tokenizer_class": "Qwen2Tokenizer",
143
+ "unk_token": null
144
+ }
training_config.yml ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ config:
2
+ (): colpali_engine.trainer.colmodel_training.ColModelTrainingConfig
3
+ output_dir: !path ../../../models/colqwen2-multi
4
+ processor:
5
+ (): colpali_engine.utils.transformers_wrappers.AllPurposeWrapper
6
+ class_to_instanciate: !ext colpali_engine.models.ColQwen2Processor
7
+ pretrained_model_name_or_path: "./models/Qwen2-VL-2B-Instruct" # "./models/paligemma-3b-mix-448"
8
+ # max_length: 50
9
+
10
+ model:
11
+ (): colpali_engine.utils.transformers_wrappers.AllPurposeWrapper
12
+ class_to_instanciate: !ext colpali_engine.models.ColQwen2
13
+ pretrained_model_name_or_path: "./models/colqwen2_base"
14
+ torch_dtype: !ext torch.bfloat16
15
+ use_cache: false
16
+ # device_map: "auto"
17
+ # quantization_config:
18
+ # (): transformers.BitsAndBytesConfig
19
+ # load_in_4bit: true
20
+ # bnb_4bit_quant_type: "nf4"
21
+ # bnb_4bit_compute_dtype: "bfloat16"
22
+ # bnb_4bit_use_double_quant: true
23
+
24
+ dataset_loading_func: !ext colpali_engine.utils.dataset_transformation.load_train_set_detailed
25
+ eval_dataset_loader: !import ../data/test_data.yaml
26
+
27
+ # max_length: 50
28
+ run_eval: true
29
+ add_suffix: true
30
+ loss_func:
31
+ (): colpali_engine.loss.late_interaction_losses.ColbertPairwiseCELoss
32
+ tr_args: !import ../tr_args/default_tr_args.yaml
33
+ peft_config:
34
+ (): peft.LoraConfig
35
+ r: 32
36
+ lora_alpha: 32
37
+ lora_dropout: 0.1
38
+ init_lora_weights: "gaussian"
39
+ bias: "none"
40
+ task_type: "FEATURE_EXTRACTION"
41
+ target_modules: '(.*(model).*(down_proj|gate_proj|up_proj|k_proj|q_proj|v_proj|o_proj).*$|.*(custom_text_proj).*$)'
42
+ # target_modules: '(.*(language_model).*(down_proj|gate_proj|up_proj|k_proj|q_proj|v_proj|o_proj).*$|.*(custom_text_proj).*$)'
43
+
vocab.json ADDED
The diff for this file is too large to render. See raw diff