PEFT
Safetensors
vidore-experimental
vidore
manu commited on
Commit
6118dbc
1 Parent(s): a91d6a6

Upload folder using huggingface_hub

Browse files
README.md ADDED
@@ -0,0 +1,202 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: HuggingFaceTB/SmolVLM-Instruct
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": "HuggingFaceTB/SmolVLM-Instruct",
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.text_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 sha256:075ad26e6ff42609469cbc3e30a907843286eef90f1527618f9aac24b9babdd7
3
+ size 72400408
added_tokens.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "<end_of_utterance>": 49154,
3
+ "<fake_token_around_image>": 49152,
4
+ "<image>": 49153
5
+ }
chat_template.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "chat_template": "<|im_start|>{% for message in messages %}{{message['role'].capitalize()}}{% if message['content'][0]['type'] == 'image' %}{{':'}}{% else %}{{': '}}{% endif %}{% for line in message['content'] %}{% if line['type'] == 'text' %}{{line['text']}}{% elif line['type'] == 'image' %}{{ '<image>' }}{% endif %}{% endfor %}<end_of_utterance>\n{% endfor %}{% if add_generation_prompt %}{{ 'Assistant:' }}{% endif %}"
3
+ }
checkpoint-2772/README.md ADDED
@@ -0,0 +1,202 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: ./models/SmolVLM-Instruct
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-2772/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "./models/SmolVLM-Instruct",
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.text_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-2772/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:075ad26e6ff42609469cbc3e30a907843286eef90f1527618f9aac24b9babdd7
3
+ size 72400408
checkpoint-2772/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d2129dcfb6562f7bc6bdad9fe51396193346aec5da9eba61402f90832898dede
3
+ size 144986754
checkpoint-2772/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a3d8db28d075b6611d668679b2d500db5ef1992227613ebb7246a23d8216db78
3
+ size 14244
checkpoint-2772/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc512ccb23c53760787387696789fcd4ddf0883dd39c3a82a3e4c7906bbec9b0
3
+ size 1064
checkpoint-2772/trainer_state.json ADDED
@@ -0,0 +1,2188 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "eval_steps": 100,
6
+ "global_step": 2772,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.010822510822510822,
13
+ "grad_norm": 1.453125,
14
+ "learning_rate": 5e-05,
15
+ "loss": 1.0068,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.021645021645021644,
20
+ "grad_norm": 0.7421875,
21
+ "learning_rate": 0.0001,
22
+ "loss": 0.7642,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.032467532467532464,
27
+ "grad_norm": 1.140625,
28
+ "learning_rate": 0.00015,
29
+ "loss": 0.676,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.04329004329004329,
34
+ "grad_norm": 0.2890625,
35
+ "learning_rate": 0.0002,
36
+ "loss": 0.6209,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.05411255411255411,
41
+ "grad_norm": 0.4453125,
42
+ "learning_rate": 0.00025,
43
+ "loss": 0.5199,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.06493506493506493,
48
+ "grad_norm": 0.353515625,
49
+ "learning_rate": 0.0003,
50
+ "loss": 0.3979,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.07575757575757576,
55
+ "grad_norm": 0.80859375,
56
+ "learning_rate": 0.00035,
57
+ "loss": 0.3112,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.08658008658008658,
62
+ "grad_norm": 0.56640625,
63
+ "learning_rate": 0.0004,
64
+ "loss": 0.2598,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.09740259740259741,
69
+ "grad_norm": 0.25,
70
+ "learning_rate": 0.00045000000000000004,
71
+ "loss": 0.2816,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.10822510822510822,
76
+ "grad_norm": 0.359375,
77
+ "learning_rate": 0.0005,
78
+ "loss": 0.2459,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.10822510822510822,
83
+ "eval_loss": 0.2404150664806366,
84
+ "eval_runtime": 62.0434,
85
+ "eval_samples_per_second": 8.059,
86
+ "eval_steps_per_second": 0.064,
87
+ "step": 100
88
+ },
89
+ {
90
+ "epoch": 0.11904761904761904,
91
+ "grad_norm": 0.447265625,
92
+ "learning_rate": 0.0004981287425149701,
93
+ "loss": 0.2341,
94
+ "step": 110
95
+ },
96
+ {
97
+ "epoch": 0.12987012987012986,
98
+ "grad_norm": 0.353515625,
99
+ "learning_rate": 0.0004962574850299401,
100
+ "loss": 0.2285,
101
+ "step": 120
102
+ },
103
+ {
104
+ "epoch": 0.1406926406926407,
105
+ "grad_norm": 0.314453125,
106
+ "learning_rate": 0.0004943862275449102,
107
+ "loss": 0.2229,
108
+ "step": 130
109
+ },
110
+ {
111
+ "epoch": 0.15151515151515152,
112
+ "grad_norm": 0.392578125,
113
+ "learning_rate": 0.0004925149700598803,
114
+ "loss": 0.2184,
115
+ "step": 140
116
+ },
117
+ {
118
+ "epoch": 0.16233766233766234,
119
+ "grad_norm": 0.263671875,
120
+ "learning_rate": 0.0004906437125748503,
121
+ "loss": 0.2225,
122
+ "step": 150
123
+ },
124
+ {
125
+ "epoch": 0.17316017316017315,
126
+ "grad_norm": 0.474609375,
127
+ "learning_rate": 0.0004887724550898204,
128
+ "loss": 0.2258,
129
+ "step": 160
130
+ },
131
+ {
132
+ "epoch": 0.18398268398268397,
133
+ "grad_norm": 0.408203125,
134
+ "learning_rate": 0.0004869011976047904,
135
+ "loss": 0.2228,
136
+ "step": 170
137
+ },
138
+ {
139
+ "epoch": 0.19480519480519481,
140
+ "grad_norm": 0.2734375,
141
+ "learning_rate": 0.00048502994011976046,
142
+ "loss": 0.2226,
143
+ "step": 180
144
+ },
145
+ {
146
+ "epoch": 0.20562770562770563,
147
+ "grad_norm": 0.236328125,
148
+ "learning_rate": 0.00048315868263473056,
149
+ "loss": 0.2105,
150
+ "step": 190
151
+ },
152
+ {
153
+ "epoch": 0.21645021645021645,
154
+ "grad_norm": 0.3125,
155
+ "learning_rate": 0.0004812874251497006,
156
+ "loss": 0.1944,
157
+ "step": 200
158
+ },
159
+ {
160
+ "epoch": 0.21645021645021645,
161
+ "eval_loss": 0.21168169379234314,
162
+ "eval_runtime": 46.9599,
163
+ "eval_samples_per_second": 10.647,
164
+ "eval_steps_per_second": 0.085,
165
+ "step": 200
166
+ },
167
+ {
168
+ "epoch": 0.22727272727272727,
169
+ "grad_norm": 0.29296875,
170
+ "learning_rate": 0.00047941616766467065,
171
+ "loss": 0.177,
172
+ "step": 210
173
+ },
174
+ {
175
+ "epoch": 0.23809523809523808,
176
+ "grad_norm": 0.203125,
177
+ "learning_rate": 0.00047754491017964074,
178
+ "loss": 0.2048,
179
+ "step": 220
180
+ },
181
+ {
182
+ "epoch": 0.24891774891774893,
183
+ "grad_norm": 0.32421875,
184
+ "learning_rate": 0.0004756736526946108,
185
+ "loss": 0.185,
186
+ "step": 230
187
+ },
188
+ {
189
+ "epoch": 0.2597402597402597,
190
+ "grad_norm": 0.279296875,
191
+ "learning_rate": 0.00047380239520958083,
192
+ "loss": 0.1941,
193
+ "step": 240
194
+ },
195
+ {
196
+ "epoch": 0.27056277056277056,
197
+ "grad_norm": 0.30078125,
198
+ "learning_rate": 0.0004719311377245509,
199
+ "loss": 0.1975,
200
+ "step": 250
201
+ },
202
+ {
203
+ "epoch": 0.2813852813852814,
204
+ "grad_norm": 0.2197265625,
205
+ "learning_rate": 0.00047005988023952097,
206
+ "loss": 0.1745,
207
+ "step": 260
208
+ },
209
+ {
210
+ "epoch": 0.2922077922077922,
211
+ "grad_norm": 0.28515625,
212
+ "learning_rate": 0.000468188622754491,
213
+ "loss": 0.1835,
214
+ "step": 270
215
+ },
216
+ {
217
+ "epoch": 0.30303030303030304,
218
+ "grad_norm": 0.2177734375,
219
+ "learning_rate": 0.0004663173652694611,
220
+ "loss": 0.2128,
221
+ "step": 280
222
+ },
223
+ {
224
+ "epoch": 0.31385281385281383,
225
+ "grad_norm": 0.2490234375,
226
+ "learning_rate": 0.00046444610778443115,
227
+ "loss": 0.1647,
228
+ "step": 290
229
+ },
230
+ {
231
+ "epoch": 0.3246753246753247,
232
+ "grad_norm": 0.2412109375,
233
+ "learning_rate": 0.0004625748502994012,
234
+ "loss": 0.1947,
235
+ "step": 300
236
+ },
237
+ {
238
+ "epoch": 0.3246753246753247,
239
+ "eval_loss": 0.18798980116844177,
240
+ "eval_runtime": 47.046,
241
+ "eval_samples_per_second": 10.628,
242
+ "eval_steps_per_second": 0.085,
243
+ "step": 300
244
+ },
245
+ {
246
+ "epoch": 0.3354978354978355,
247
+ "grad_norm": 0.2392578125,
248
+ "learning_rate": 0.00046070359281437124,
249
+ "loss": 0.2069,
250
+ "step": 310
251
+ },
252
+ {
253
+ "epoch": 0.3463203463203463,
254
+ "grad_norm": 0.26953125,
255
+ "learning_rate": 0.00045883233532934134,
256
+ "loss": 0.2029,
257
+ "step": 320
258
+ },
259
+ {
260
+ "epoch": 0.35714285714285715,
261
+ "grad_norm": 0.26171875,
262
+ "learning_rate": 0.0004569610778443114,
263
+ "loss": 0.2126,
264
+ "step": 330
265
+ },
266
+ {
267
+ "epoch": 0.36796536796536794,
268
+ "grad_norm": 0.16015625,
269
+ "learning_rate": 0.0004550898203592814,
270
+ "loss": 0.1838,
271
+ "step": 340
272
+ },
273
+ {
274
+ "epoch": 0.3787878787878788,
275
+ "grad_norm": 0.271484375,
276
+ "learning_rate": 0.0004532185628742515,
277
+ "loss": 0.1869,
278
+ "step": 350
279
+ },
280
+ {
281
+ "epoch": 0.38961038961038963,
282
+ "grad_norm": 0.2333984375,
283
+ "learning_rate": 0.00045134730538922156,
284
+ "loss": 0.1897,
285
+ "step": 360
286
+ },
287
+ {
288
+ "epoch": 0.4004329004329004,
289
+ "grad_norm": 0.2236328125,
290
+ "learning_rate": 0.0004494760479041916,
291
+ "loss": 0.1896,
292
+ "step": 370
293
+ },
294
+ {
295
+ "epoch": 0.41125541125541126,
296
+ "grad_norm": 0.275390625,
297
+ "learning_rate": 0.0004476047904191617,
298
+ "loss": 0.1956,
299
+ "step": 380
300
+ },
301
+ {
302
+ "epoch": 0.42207792207792205,
303
+ "grad_norm": 0.2099609375,
304
+ "learning_rate": 0.00044573353293413174,
305
+ "loss": 0.1776,
306
+ "step": 390
307
+ },
308
+ {
309
+ "epoch": 0.4329004329004329,
310
+ "grad_norm": 0.2314453125,
311
+ "learning_rate": 0.0004438622754491018,
312
+ "loss": 0.1866,
313
+ "step": 400
314
+ },
315
+ {
316
+ "epoch": 0.4329004329004329,
317
+ "eval_loss": 0.1857587993144989,
318
+ "eval_runtime": 49.3784,
319
+ "eval_samples_per_second": 10.126,
320
+ "eval_steps_per_second": 0.081,
321
+ "step": 400
322
+ },
323
+ {
324
+ "epoch": 0.44372294372294374,
325
+ "grad_norm": 0.37109375,
326
+ "learning_rate": 0.0004419910179640719,
327
+ "loss": 0.1785,
328
+ "step": 410
329
+ },
330
+ {
331
+ "epoch": 0.45454545454545453,
332
+ "grad_norm": 0.26953125,
333
+ "learning_rate": 0.00044011976047904193,
334
+ "loss": 0.2015,
335
+ "step": 420
336
+ },
337
+ {
338
+ "epoch": 0.4653679653679654,
339
+ "grad_norm": 0.26171875,
340
+ "learning_rate": 0.00043824850299401197,
341
+ "loss": 0.1682,
342
+ "step": 430
343
+ },
344
+ {
345
+ "epoch": 0.47619047619047616,
346
+ "grad_norm": 0.2734375,
347
+ "learning_rate": 0.00043637724550898207,
348
+ "loss": 0.1783,
349
+ "step": 440
350
+ },
351
+ {
352
+ "epoch": 0.487012987012987,
353
+ "grad_norm": 0.2138671875,
354
+ "learning_rate": 0.00043450598802395206,
355
+ "loss": 0.188,
356
+ "step": 450
357
+ },
358
+ {
359
+ "epoch": 0.49783549783549785,
360
+ "grad_norm": 0.1669921875,
361
+ "learning_rate": 0.00043263473053892215,
362
+ "loss": 0.1804,
363
+ "step": 460
364
+ },
365
+ {
366
+ "epoch": 0.5086580086580087,
367
+ "grad_norm": 0.26953125,
368
+ "learning_rate": 0.00043076347305389225,
369
+ "loss": 0.1698,
370
+ "step": 470
371
+ },
372
+ {
373
+ "epoch": 0.5194805194805194,
374
+ "grad_norm": 0.291015625,
375
+ "learning_rate": 0.00042889221556886224,
376
+ "loss": 0.1776,
377
+ "step": 480
378
+ },
379
+ {
380
+ "epoch": 0.5303030303030303,
381
+ "grad_norm": 0.287109375,
382
+ "learning_rate": 0.00042702095808383234,
383
+ "loss": 0.1753,
384
+ "step": 490
385
+ },
386
+ {
387
+ "epoch": 0.5411255411255411,
388
+ "grad_norm": 0.244140625,
389
+ "learning_rate": 0.00042514970059880243,
390
+ "loss": 0.1786,
391
+ "step": 500
392
+ },
393
+ {
394
+ "epoch": 0.5411255411255411,
395
+ "eval_loss": 0.18613742291927338,
396
+ "eval_runtime": 49.4353,
397
+ "eval_samples_per_second": 10.114,
398
+ "eval_steps_per_second": 0.081,
399
+ "step": 500
400
+ },
401
+ {
402
+ "epoch": 0.551948051948052,
403
+ "grad_norm": 0.30859375,
404
+ "learning_rate": 0.0004232784431137724,
405
+ "loss": 0.1558,
406
+ "step": 510
407
+ },
408
+ {
409
+ "epoch": 0.5627705627705628,
410
+ "grad_norm": 0.1484375,
411
+ "learning_rate": 0.0004214071856287425,
412
+ "loss": 0.1595,
413
+ "step": 520
414
+ },
415
+ {
416
+ "epoch": 0.5735930735930735,
417
+ "grad_norm": 0.208984375,
418
+ "learning_rate": 0.0004195359281437126,
419
+ "loss": 0.161,
420
+ "step": 530
421
+ },
422
+ {
423
+ "epoch": 0.5844155844155844,
424
+ "grad_norm": 0.21484375,
425
+ "learning_rate": 0.00041766467065868266,
426
+ "loss": 0.1699,
427
+ "step": 540
428
+ },
429
+ {
430
+ "epoch": 0.5952380952380952,
431
+ "grad_norm": 0.2890625,
432
+ "learning_rate": 0.0004157934131736527,
433
+ "loss": 0.1694,
434
+ "step": 550
435
+ },
436
+ {
437
+ "epoch": 0.6060606060606061,
438
+ "grad_norm": 0.1943359375,
439
+ "learning_rate": 0.00041392215568862275,
440
+ "loss": 0.1753,
441
+ "step": 560
442
+ },
443
+ {
444
+ "epoch": 0.6168831168831169,
445
+ "grad_norm": 0.185546875,
446
+ "learning_rate": 0.00041205089820359284,
447
+ "loss": 0.1781,
448
+ "step": 570
449
+ },
450
+ {
451
+ "epoch": 0.6277056277056277,
452
+ "grad_norm": 0.283203125,
453
+ "learning_rate": 0.0004101796407185629,
454
+ "loss": 0.1778,
455
+ "step": 580
456
+ },
457
+ {
458
+ "epoch": 0.6385281385281385,
459
+ "grad_norm": 0.208984375,
460
+ "learning_rate": 0.00040830838323353293,
461
+ "loss": 0.1501,
462
+ "step": 590
463
+ },
464
+ {
465
+ "epoch": 0.6493506493506493,
466
+ "grad_norm": 0.2021484375,
467
+ "learning_rate": 0.000406437125748503,
468
+ "loss": 0.1855,
469
+ "step": 600
470
+ },
471
+ {
472
+ "epoch": 0.6493506493506493,
473
+ "eval_loss": 0.18048742413520813,
474
+ "eval_runtime": 52.6384,
475
+ "eval_samples_per_second": 9.499,
476
+ "eval_steps_per_second": 0.076,
477
+ "step": 600
478
+ },
479
+ {
480
+ "epoch": 0.6601731601731602,
481
+ "grad_norm": 0.162109375,
482
+ "learning_rate": 0.00040456586826347307,
483
+ "loss": 0.1783,
484
+ "step": 610
485
+ },
486
+ {
487
+ "epoch": 0.670995670995671,
488
+ "grad_norm": 0.330078125,
489
+ "learning_rate": 0.0004026946107784431,
490
+ "loss": 0.1604,
491
+ "step": 620
492
+ },
493
+ {
494
+ "epoch": 0.6818181818181818,
495
+ "grad_norm": 0.205078125,
496
+ "learning_rate": 0.0004008233532934132,
497
+ "loss": 0.168,
498
+ "step": 630
499
+ },
500
+ {
501
+ "epoch": 0.6926406926406926,
502
+ "grad_norm": 0.2080078125,
503
+ "learning_rate": 0.00039895209580838325,
504
+ "loss": 0.1583,
505
+ "step": 640
506
+ },
507
+ {
508
+ "epoch": 0.7034632034632035,
509
+ "grad_norm": 0.203125,
510
+ "learning_rate": 0.0003970808383233533,
511
+ "loss": 0.1432,
512
+ "step": 650
513
+ },
514
+ {
515
+ "epoch": 0.7142857142857143,
516
+ "grad_norm": 0.2177734375,
517
+ "learning_rate": 0.0003952095808383234,
518
+ "loss": 0.1823,
519
+ "step": 660
520
+ },
521
+ {
522
+ "epoch": 0.7251082251082251,
523
+ "grad_norm": 0.2275390625,
524
+ "learning_rate": 0.00039333832335329343,
525
+ "loss": 0.1776,
526
+ "step": 670
527
+ },
528
+ {
529
+ "epoch": 0.7359307359307359,
530
+ "grad_norm": 0.234375,
531
+ "learning_rate": 0.0003914670658682635,
532
+ "loss": 0.1513,
533
+ "step": 680
534
+ },
535
+ {
536
+ "epoch": 0.7467532467532467,
537
+ "grad_norm": 0.283203125,
538
+ "learning_rate": 0.0003895958083832336,
539
+ "loss": 0.1692,
540
+ "step": 690
541
+ },
542
+ {
543
+ "epoch": 0.7575757575757576,
544
+ "grad_norm": 0.140625,
545
+ "learning_rate": 0.00038772455089820356,
546
+ "loss": 0.1716,
547
+ "step": 700
548
+ },
549
+ {
550
+ "epoch": 0.7575757575757576,
551
+ "eval_loss": 0.16686995327472687,
552
+ "eval_runtime": 48.5544,
553
+ "eval_samples_per_second": 10.298,
554
+ "eval_steps_per_second": 0.082,
555
+ "step": 700
556
+ },
557
+ {
558
+ "epoch": 0.7683982683982684,
559
+ "grad_norm": 0.28125,
560
+ "learning_rate": 0.00038585329341317366,
561
+ "loss": 0.1619,
562
+ "step": 710
563
+ },
564
+ {
565
+ "epoch": 0.7792207792207793,
566
+ "grad_norm": 0.2197265625,
567
+ "learning_rate": 0.00038398203592814376,
568
+ "loss": 0.167,
569
+ "step": 720
570
+ },
571
+ {
572
+ "epoch": 0.79004329004329,
573
+ "grad_norm": 0.3125,
574
+ "learning_rate": 0.00038211077844311375,
575
+ "loss": 0.1507,
576
+ "step": 730
577
+ },
578
+ {
579
+ "epoch": 0.8008658008658008,
580
+ "grad_norm": 0.390625,
581
+ "learning_rate": 0.00038023952095808384,
582
+ "loss": 0.1649,
583
+ "step": 740
584
+ },
585
+ {
586
+ "epoch": 0.8116883116883117,
587
+ "grad_norm": 0.185546875,
588
+ "learning_rate": 0.00037836826347305394,
589
+ "loss": 0.1537,
590
+ "step": 750
591
+ },
592
+ {
593
+ "epoch": 0.8225108225108225,
594
+ "grad_norm": 0.2021484375,
595
+ "learning_rate": 0.00037649700598802393,
596
+ "loss": 0.1617,
597
+ "step": 760
598
+ },
599
+ {
600
+ "epoch": 0.8333333333333334,
601
+ "grad_norm": 0.267578125,
602
+ "learning_rate": 0.00037462574850299403,
603
+ "loss": 0.1641,
604
+ "step": 770
605
+ },
606
+ {
607
+ "epoch": 0.8441558441558441,
608
+ "grad_norm": 0.296875,
609
+ "learning_rate": 0.0003727544910179641,
610
+ "loss": 0.1742,
611
+ "step": 780
612
+ },
613
+ {
614
+ "epoch": 0.854978354978355,
615
+ "grad_norm": 0.2265625,
616
+ "learning_rate": 0.0003708832335329341,
617
+ "loss": 0.1787,
618
+ "step": 790
619
+ },
620
+ {
621
+ "epoch": 0.8658008658008658,
622
+ "grad_norm": 0.1845703125,
623
+ "learning_rate": 0.0003690119760479042,
624
+ "loss": 0.1736,
625
+ "step": 800
626
+ },
627
+ {
628
+ "epoch": 0.8658008658008658,
629
+ "eval_loss": 0.16875208914279938,
630
+ "eval_runtime": 49.7255,
631
+ "eval_samples_per_second": 10.055,
632
+ "eval_steps_per_second": 0.08,
633
+ "step": 800
634
+ },
635
+ {
636
+ "epoch": 0.8766233766233766,
637
+ "grad_norm": 0.2109375,
638
+ "learning_rate": 0.00036714071856287425,
639
+ "loss": 0.1729,
640
+ "step": 810
641
+ },
642
+ {
643
+ "epoch": 0.8874458874458875,
644
+ "grad_norm": 0.392578125,
645
+ "learning_rate": 0.0003652694610778443,
646
+ "loss": 0.1617,
647
+ "step": 820
648
+ },
649
+ {
650
+ "epoch": 0.8982683982683982,
651
+ "grad_norm": 0.2578125,
652
+ "learning_rate": 0.0003633982035928144,
653
+ "loss": 0.1637,
654
+ "step": 830
655
+ },
656
+ {
657
+ "epoch": 0.9090909090909091,
658
+ "grad_norm": 0.2109375,
659
+ "learning_rate": 0.00036152694610778444,
660
+ "loss": 0.1569,
661
+ "step": 840
662
+ },
663
+ {
664
+ "epoch": 0.9199134199134199,
665
+ "grad_norm": 0.18359375,
666
+ "learning_rate": 0.0003596556886227545,
667
+ "loss": 0.1547,
668
+ "step": 850
669
+ },
670
+ {
671
+ "epoch": 0.9307359307359307,
672
+ "grad_norm": 0.2197265625,
673
+ "learning_rate": 0.0003577844311377246,
674
+ "loss": 0.1645,
675
+ "step": 860
676
+ },
677
+ {
678
+ "epoch": 0.9415584415584416,
679
+ "grad_norm": 0.357421875,
680
+ "learning_rate": 0.0003559131736526946,
681
+ "loss": 0.163,
682
+ "step": 870
683
+ },
684
+ {
685
+ "epoch": 0.9523809523809523,
686
+ "grad_norm": 0.255859375,
687
+ "learning_rate": 0.00035404191616766466,
688
+ "loss": 0.1748,
689
+ "step": 880
690
+ },
691
+ {
692
+ "epoch": 0.9632034632034632,
693
+ "grad_norm": 0.19921875,
694
+ "learning_rate": 0.00035217065868263476,
695
+ "loss": 0.1672,
696
+ "step": 890
697
+ },
698
+ {
699
+ "epoch": 0.974025974025974,
700
+ "grad_norm": 0.330078125,
701
+ "learning_rate": 0.0003502994011976048,
702
+ "loss": 0.1699,
703
+ "step": 900
704
+ },
705
+ {
706
+ "epoch": 0.974025974025974,
707
+ "eval_loss": 0.16794857382774353,
708
+ "eval_runtime": 58.792,
709
+ "eval_samples_per_second": 8.505,
710
+ "eval_steps_per_second": 0.068,
711
+ "step": 900
712
+ },
713
+ {
714
+ "epoch": 0.9848484848484849,
715
+ "grad_norm": 0.1435546875,
716
+ "learning_rate": 0.00034842814371257485,
717
+ "loss": 0.1537,
718
+ "step": 910
719
+ },
720
+ {
721
+ "epoch": 0.9956709956709957,
722
+ "grad_norm": 0.2421875,
723
+ "learning_rate": 0.00034655688622754494,
724
+ "loss": 0.1629,
725
+ "step": 920
726
+ },
727
+ {
728
+ "epoch": 1.0064935064935066,
729
+ "grad_norm": 0.19140625,
730
+ "learning_rate": 0.000344685628742515,
731
+ "loss": 0.1499,
732
+ "step": 930
733
+ },
734
+ {
735
+ "epoch": 1.0173160173160174,
736
+ "grad_norm": 0.28125,
737
+ "learning_rate": 0.00034281437125748503,
738
+ "loss": 0.1464,
739
+ "step": 940
740
+ },
741
+ {
742
+ "epoch": 1.0281385281385282,
743
+ "grad_norm": 0.173828125,
744
+ "learning_rate": 0.00034094311377245507,
745
+ "loss": 0.1416,
746
+ "step": 950
747
+ },
748
+ {
749
+ "epoch": 1.0389610389610389,
750
+ "grad_norm": 0.2119140625,
751
+ "learning_rate": 0.00033907185628742517,
752
+ "loss": 0.1545,
753
+ "step": 960
754
+ },
755
+ {
756
+ "epoch": 1.0497835497835497,
757
+ "grad_norm": 0.18359375,
758
+ "learning_rate": 0.0003372005988023952,
759
+ "loss": 0.1431,
760
+ "step": 970
761
+ },
762
+ {
763
+ "epoch": 1.0606060606060606,
764
+ "grad_norm": 0.279296875,
765
+ "learning_rate": 0.00033532934131736525,
766
+ "loss": 0.1607,
767
+ "step": 980
768
+ },
769
+ {
770
+ "epoch": 1.0714285714285714,
771
+ "grad_norm": 0.384765625,
772
+ "learning_rate": 0.00033345808383233535,
773
+ "loss": 0.1602,
774
+ "step": 990
775
+ },
776
+ {
777
+ "epoch": 1.0822510822510822,
778
+ "grad_norm": 0.2470703125,
779
+ "learning_rate": 0.0003315868263473054,
780
+ "loss": 0.1651,
781
+ "step": 1000
782
+ },
783
+ {
784
+ "epoch": 1.0822510822510822,
785
+ "eval_loss": 0.1569806933403015,
786
+ "eval_runtime": 57.763,
787
+ "eval_samples_per_second": 8.656,
788
+ "eval_steps_per_second": 0.069,
789
+ "step": 1000
790
+ },
791
+ {
792
+ "epoch": 1.093073593073593,
793
+ "grad_norm": 0.2734375,
794
+ "learning_rate": 0.00032971556886227544,
795
+ "loss": 0.146,
796
+ "step": 1010
797
+ },
798
+ {
799
+ "epoch": 1.103896103896104,
800
+ "grad_norm": 0.275390625,
801
+ "learning_rate": 0.00032784431137724553,
802
+ "loss": 0.1341,
803
+ "step": 1020
804
+ },
805
+ {
806
+ "epoch": 1.1147186147186148,
807
+ "grad_norm": 0.1640625,
808
+ "learning_rate": 0.0003259730538922156,
809
+ "loss": 0.1412,
810
+ "step": 1030
811
+ },
812
+ {
813
+ "epoch": 1.1255411255411256,
814
+ "grad_norm": 0.21875,
815
+ "learning_rate": 0.0003241017964071856,
816
+ "loss": 0.152,
817
+ "step": 1040
818
+ },
819
+ {
820
+ "epoch": 1.1363636363636362,
821
+ "grad_norm": 0.1962890625,
822
+ "learning_rate": 0.0003222305389221557,
823
+ "loss": 0.1501,
824
+ "step": 1050
825
+ },
826
+ {
827
+ "epoch": 1.1471861471861473,
828
+ "grad_norm": 0.1337890625,
829
+ "learning_rate": 0.00032035928143712576,
830
+ "loss": 0.1359,
831
+ "step": 1060
832
+ },
833
+ {
834
+ "epoch": 1.158008658008658,
835
+ "grad_norm": 0.271484375,
836
+ "learning_rate": 0.0003184880239520958,
837
+ "loss": 0.1437,
838
+ "step": 1070
839
+ },
840
+ {
841
+ "epoch": 1.1688311688311688,
842
+ "grad_norm": 0.2216796875,
843
+ "learning_rate": 0.0003166167664670659,
844
+ "loss": 0.1395,
845
+ "step": 1080
846
+ },
847
+ {
848
+ "epoch": 1.1796536796536796,
849
+ "grad_norm": 0.314453125,
850
+ "learning_rate": 0.0003147455089820359,
851
+ "loss": 0.1403,
852
+ "step": 1090
853
+ },
854
+ {
855
+ "epoch": 1.1904761904761905,
856
+ "grad_norm": 0.2236328125,
857
+ "learning_rate": 0.000312874251497006,
858
+ "loss": 0.157,
859
+ "step": 1100
860
+ },
861
+ {
862
+ "epoch": 1.1904761904761905,
863
+ "eval_loss": 0.15362851321697235,
864
+ "eval_runtime": 50.1243,
865
+ "eval_samples_per_second": 9.975,
866
+ "eval_steps_per_second": 0.08,
867
+ "step": 1100
868
+ },
869
+ {
870
+ "epoch": 1.2012987012987013,
871
+ "grad_norm": 0.2177734375,
872
+ "learning_rate": 0.0003110029940119761,
873
+ "loss": 0.1377,
874
+ "step": 1110
875
+ },
876
+ {
877
+ "epoch": 1.2121212121212122,
878
+ "grad_norm": 0.2138671875,
879
+ "learning_rate": 0.0003091317365269461,
880
+ "loss": 0.1423,
881
+ "step": 1120
882
+ },
883
+ {
884
+ "epoch": 1.222943722943723,
885
+ "grad_norm": 0.1953125,
886
+ "learning_rate": 0.00030726047904191617,
887
+ "loss": 0.1353,
888
+ "step": 1130
889
+ },
890
+ {
891
+ "epoch": 1.2337662337662338,
892
+ "grad_norm": 0.255859375,
893
+ "learning_rate": 0.00030538922155688627,
894
+ "loss": 0.1453,
895
+ "step": 1140
896
+ },
897
+ {
898
+ "epoch": 1.2445887445887447,
899
+ "grad_norm": 0.3828125,
900
+ "learning_rate": 0.00030351796407185626,
901
+ "loss": 0.1398,
902
+ "step": 1150
903
+ },
904
+ {
905
+ "epoch": 1.2554112554112553,
906
+ "grad_norm": 0.189453125,
907
+ "learning_rate": 0.00030164670658682635,
908
+ "loss": 0.1315,
909
+ "step": 1160
910
+ },
911
+ {
912
+ "epoch": 1.2662337662337662,
913
+ "grad_norm": 0.26953125,
914
+ "learning_rate": 0.00029977544910179645,
915
+ "loss": 0.1323,
916
+ "step": 1170
917
+ },
918
+ {
919
+ "epoch": 1.277056277056277,
920
+ "grad_norm": 0.1513671875,
921
+ "learning_rate": 0.00029790419161676644,
922
+ "loss": 0.134,
923
+ "step": 1180
924
+ },
925
+ {
926
+ "epoch": 1.2878787878787878,
927
+ "grad_norm": 0.3046875,
928
+ "learning_rate": 0.00029603293413173654,
929
+ "loss": 0.1544,
930
+ "step": 1190
931
+ },
932
+ {
933
+ "epoch": 1.2987012987012987,
934
+ "grad_norm": 0.236328125,
935
+ "learning_rate": 0.0002941616766467066,
936
+ "loss": 0.141,
937
+ "step": 1200
938
+ },
939
+ {
940
+ "epoch": 1.2987012987012987,
941
+ "eval_loss": 0.1604161411523819,
942
+ "eval_runtime": 51.9642,
943
+ "eval_samples_per_second": 9.622,
944
+ "eval_steps_per_second": 0.077,
945
+ "step": 1200
946
+ },
947
+ {
948
+ "epoch": 1.3095238095238095,
949
+ "grad_norm": 0.14453125,
950
+ "learning_rate": 0.0002922904191616766,
951
+ "loss": 0.1405,
952
+ "step": 1210
953
+ },
954
+ {
955
+ "epoch": 1.3203463203463204,
956
+ "grad_norm": 0.154296875,
957
+ "learning_rate": 0.0002904191616766467,
958
+ "loss": 0.1388,
959
+ "step": 1220
960
+ },
961
+ {
962
+ "epoch": 1.3311688311688312,
963
+ "grad_norm": 0.1572265625,
964
+ "learning_rate": 0.00028854790419161676,
965
+ "loss": 0.1482,
966
+ "step": 1230
967
+ },
968
+ {
969
+ "epoch": 1.341991341991342,
970
+ "grad_norm": 0.201171875,
971
+ "learning_rate": 0.0002866766467065868,
972
+ "loss": 0.1299,
973
+ "step": 1240
974
+ },
975
+ {
976
+ "epoch": 1.3528138528138527,
977
+ "grad_norm": 0.2265625,
978
+ "learning_rate": 0.0002848053892215569,
979
+ "loss": 0.137,
980
+ "step": 1250
981
+ },
982
+ {
983
+ "epoch": 1.3636363636363638,
984
+ "grad_norm": 0.2255859375,
985
+ "learning_rate": 0.00028293413173652695,
986
+ "loss": 0.1173,
987
+ "step": 1260
988
+ },
989
+ {
990
+ "epoch": 1.3744588744588744,
991
+ "grad_norm": 0.25,
992
+ "learning_rate": 0.000281062874251497,
993
+ "loss": 0.1317,
994
+ "step": 1270
995
+ },
996
+ {
997
+ "epoch": 1.3852813852813852,
998
+ "grad_norm": 0.1455078125,
999
+ "learning_rate": 0.0002791916167664671,
1000
+ "loss": 0.1475,
1001
+ "step": 1280
1002
+ },
1003
+ {
1004
+ "epoch": 1.396103896103896,
1005
+ "grad_norm": 0.150390625,
1006
+ "learning_rate": 0.00027732035928143713,
1007
+ "loss": 0.1411,
1008
+ "step": 1290
1009
+ },
1010
+ {
1011
+ "epoch": 1.406926406926407,
1012
+ "grad_norm": 0.1552734375,
1013
+ "learning_rate": 0.00027544910179640717,
1014
+ "loss": 0.1343,
1015
+ "step": 1300
1016
+ },
1017
+ {
1018
+ "epoch": 1.406926406926407,
1019
+ "eval_loss": 0.1617971956729889,
1020
+ "eval_runtime": 53.8477,
1021
+ "eval_samples_per_second": 9.285,
1022
+ "eval_steps_per_second": 0.074,
1023
+ "step": 1300
1024
+ },
1025
+ {
1026
+ "epoch": 1.4177489177489178,
1027
+ "grad_norm": 0.26171875,
1028
+ "learning_rate": 0.00027357784431137727,
1029
+ "loss": 0.1367,
1030
+ "step": 1310
1031
+ },
1032
+ {
1033
+ "epoch": 1.4285714285714286,
1034
+ "grad_norm": 0.154296875,
1035
+ "learning_rate": 0.0002717065868263473,
1036
+ "loss": 0.1359,
1037
+ "step": 1320
1038
+ },
1039
+ {
1040
+ "epoch": 1.4393939393939394,
1041
+ "grad_norm": 0.208984375,
1042
+ "learning_rate": 0.00026983532934131735,
1043
+ "loss": 0.1257,
1044
+ "step": 1330
1045
+ },
1046
+ {
1047
+ "epoch": 1.4502164502164503,
1048
+ "grad_norm": 0.1865234375,
1049
+ "learning_rate": 0.0002679640718562874,
1050
+ "loss": 0.1282,
1051
+ "step": 1340
1052
+ },
1053
+ {
1054
+ "epoch": 1.4610389610389611,
1055
+ "grad_norm": 0.1767578125,
1056
+ "learning_rate": 0.0002660928143712575,
1057
+ "loss": 0.1266,
1058
+ "step": 1350
1059
+ },
1060
+ {
1061
+ "epoch": 1.4718614718614718,
1062
+ "grad_norm": 0.1982421875,
1063
+ "learning_rate": 0.0002642215568862276,
1064
+ "loss": 0.1204,
1065
+ "step": 1360
1066
+ },
1067
+ {
1068
+ "epoch": 1.4826839826839826,
1069
+ "grad_norm": 0.1962890625,
1070
+ "learning_rate": 0.0002623502994011976,
1071
+ "loss": 0.151,
1072
+ "step": 1370
1073
+ },
1074
+ {
1075
+ "epoch": 1.4935064935064934,
1076
+ "grad_norm": 0.20703125,
1077
+ "learning_rate": 0.0002604790419161677,
1078
+ "loss": 0.1336,
1079
+ "step": 1380
1080
+ },
1081
+ {
1082
+ "epoch": 1.5043290043290043,
1083
+ "grad_norm": 0.1376953125,
1084
+ "learning_rate": 0.0002586077844311378,
1085
+ "loss": 0.1364,
1086
+ "step": 1390
1087
+ },
1088
+ {
1089
+ "epoch": 1.5151515151515151,
1090
+ "grad_norm": 0.333984375,
1091
+ "learning_rate": 0.00025673652694610776,
1092
+ "loss": 0.1466,
1093
+ "step": 1400
1094
+ },
1095
+ {
1096
+ "epoch": 1.5151515151515151,
1097
+ "eval_loss": 0.165997713804245,
1098
+ "eval_runtime": 49.4239,
1099
+ "eval_samples_per_second": 10.117,
1100
+ "eval_steps_per_second": 0.081,
1101
+ "step": 1400
1102
+ },
1103
+ {
1104
+ "epoch": 1.525974025974026,
1105
+ "grad_norm": 0.154296875,
1106
+ "learning_rate": 0.00025486526946107786,
1107
+ "loss": 0.1306,
1108
+ "step": 1410
1109
+ },
1110
+ {
1111
+ "epoch": 1.5367965367965368,
1112
+ "grad_norm": 0.1630859375,
1113
+ "learning_rate": 0.00025299401197604796,
1114
+ "loss": 0.1246,
1115
+ "step": 1420
1116
+ },
1117
+ {
1118
+ "epoch": 1.5476190476190477,
1119
+ "grad_norm": 0.1904296875,
1120
+ "learning_rate": 0.00025112275449101795,
1121
+ "loss": 0.1333,
1122
+ "step": 1430
1123
+ },
1124
+ {
1125
+ "epoch": 1.5584415584415585,
1126
+ "grad_norm": 0.201171875,
1127
+ "learning_rate": 0.00024925149700598804,
1128
+ "loss": 0.1258,
1129
+ "step": 1440
1130
+ },
1131
+ {
1132
+ "epoch": 1.5692640692640691,
1133
+ "grad_norm": 0.244140625,
1134
+ "learning_rate": 0.0002473802395209581,
1135
+ "loss": 0.1493,
1136
+ "step": 1450
1137
+ },
1138
+ {
1139
+ "epoch": 1.5800865800865802,
1140
+ "grad_norm": 0.1279296875,
1141
+ "learning_rate": 0.00024550898203592813,
1142
+ "loss": 0.1346,
1143
+ "step": 1460
1144
+ },
1145
+ {
1146
+ "epoch": 1.5909090909090908,
1147
+ "grad_norm": 0.154296875,
1148
+ "learning_rate": 0.00024363772455089823,
1149
+ "loss": 0.1228,
1150
+ "step": 1470
1151
+ },
1152
+ {
1153
+ "epoch": 1.601731601731602,
1154
+ "grad_norm": 0.22265625,
1155
+ "learning_rate": 0.00024176646706586827,
1156
+ "loss": 0.1335,
1157
+ "step": 1480
1158
+ },
1159
+ {
1160
+ "epoch": 1.6125541125541125,
1161
+ "grad_norm": 0.2333984375,
1162
+ "learning_rate": 0.0002398952095808383,
1163
+ "loss": 0.137,
1164
+ "step": 1490
1165
+ },
1166
+ {
1167
+ "epoch": 1.6233766233766234,
1168
+ "grad_norm": 0.197265625,
1169
+ "learning_rate": 0.00023802395209580838,
1170
+ "loss": 0.1351,
1171
+ "step": 1500
1172
+ },
1173
+ {
1174
+ "epoch": 1.6233766233766234,
1175
+ "eval_loss": 0.1522991955280304,
1176
+ "eval_runtime": 49.2679,
1177
+ "eval_samples_per_second": 10.149,
1178
+ "eval_steps_per_second": 0.081,
1179
+ "step": 1500
1180
+ },
1181
+ {
1182
+ "epoch": 1.6341991341991342,
1183
+ "grad_norm": 0.330078125,
1184
+ "learning_rate": 0.00023615269461077845,
1185
+ "loss": 0.1361,
1186
+ "step": 1510
1187
+ },
1188
+ {
1189
+ "epoch": 1.645021645021645,
1190
+ "grad_norm": 0.140625,
1191
+ "learning_rate": 0.0002342814371257485,
1192
+ "loss": 0.1383,
1193
+ "step": 1520
1194
+ },
1195
+ {
1196
+ "epoch": 1.655844155844156,
1197
+ "grad_norm": 0.1826171875,
1198
+ "learning_rate": 0.00023241017964071857,
1199
+ "loss": 0.135,
1200
+ "step": 1530
1201
+ },
1202
+ {
1203
+ "epoch": 1.6666666666666665,
1204
+ "grad_norm": 0.2197265625,
1205
+ "learning_rate": 0.00023053892215568864,
1206
+ "loss": 0.1298,
1207
+ "step": 1540
1208
+ },
1209
+ {
1210
+ "epoch": 1.6774891774891776,
1211
+ "grad_norm": 0.2080078125,
1212
+ "learning_rate": 0.00022866766467065868,
1213
+ "loss": 0.1323,
1214
+ "step": 1550
1215
+ },
1216
+ {
1217
+ "epoch": 1.6883116883116882,
1218
+ "grad_norm": 0.232421875,
1219
+ "learning_rate": 0.00022679640718562875,
1220
+ "loss": 0.1253,
1221
+ "step": 1560
1222
+ },
1223
+ {
1224
+ "epoch": 1.6991341991341993,
1225
+ "grad_norm": 0.130859375,
1226
+ "learning_rate": 0.0002249251497005988,
1227
+ "loss": 0.1324,
1228
+ "step": 1570
1229
+ },
1230
+ {
1231
+ "epoch": 1.70995670995671,
1232
+ "grad_norm": 0.1865234375,
1233
+ "learning_rate": 0.0002230538922155689,
1234
+ "loss": 0.1402,
1235
+ "step": 1580
1236
+ },
1237
+ {
1238
+ "epoch": 1.7207792207792207,
1239
+ "grad_norm": 0.1875,
1240
+ "learning_rate": 0.00022118263473053893,
1241
+ "loss": 0.1307,
1242
+ "step": 1590
1243
+ },
1244
+ {
1245
+ "epoch": 1.7316017316017316,
1246
+ "grad_norm": 0.12890625,
1247
+ "learning_rate": 0.00021931137724550898,
1248
+ "loss": 0.1276,
1249
+ "step": 1600
1250
+ },
1251
+ {
1252
+ "epoch": 1.7316017316017316,
1253
+ "eval_loss": 0.1606496274471283,
1254
+ "eval_runtime": 51.8902,
1255
+ "eval_samples_per_second": 9.636,
1256
+ "eval_steps_per_second": 0.077,
1257
+ "step": 1600
1258
+ },
1259
+ {
1260
+ "epoch": 1.7424242424242424,
1261
+ "grad_norm": 0.1455078125,
1262
+ "learning_rate": 0.00021744011976047905,
1263
+ "loss": 0.1302,
1264
+ "step": 1610
1265
+ },
1266
+ {
1267
+ "epoch": 1.7532467532467533,
1268
+ "grad_norm": 0.1455078125,
1269
+ "learning_rate": 0.00021556886227544912,
1270
+ "loss": 0.1231,
1271
+ "step": 1620
1272
+ },
1273
+ {
1274
+ "epoch": 1.7640692640692641,
1275
+ "grad_norm": 0.162109375,
1276
+ "learning_rate": 0.00021369760479041916,
1277
+ "loss": 0.1362,
1278
+ "step": 1630
1279
+ },
1280
+ {
1281
+ "epoch": 1.774891774891775,
1282
+ "grad_norm": 0.16796875,
1283
+ "learning_rate": 0.00021182634730538923,
1284
+ "loss": 0.1264,
1285
+ "step": 1640
1286
+ },
1287
+ {
1288
+ "epoch": 1.7857142857142856,
1289
+ "grad_norm": 0.25390625,
1290
+ "learning_rate": 0.0002099550898203593,
1291
+ "loss": 0.142,
1292
+ "step": 1650
1293
+ },
1294
+ {
1295
+ "epoch": 1.7965367965367967,
1296
+ "grad_norm": 0.1279296875,
1297
+ "learning_rate": 0.00020808383233532934,
1298
+ "loss": 0.1347,
1299
+ "step": 1660
1300
+ },
1301
+ {
1302
+ "epoch": 1.8073593073593073,
1303
+ "grad_norm": 0.1708984375,
1304
+ "learning_rate": 0.0002062125748502994,
1305
+ "loss": 0.1399,
1306
+ "step": 1670
1307
+ },
1308
+ {
1309
+ "epoch": 1.8181818181818183,
1310
+ "grad_norm": 0.24609375,
1311
+ "learning_rate": 0.00020434131736526945,
1312
+ "loss": 0.1236,
1313
+ "step": 1680
1314
+ },
1315
+ {
1316
+ "epoch": 1.829004329004329,
1317
+ "grad_norm": 0.2255859375,
1318
+ "learning_rate": 0.00020247005988023952,
1319
+ "loss": 0.113,
1320
+ "step": 1690
1321
+ },
1322
+ {
1323
+ "epoch": 1.8398268398268398,
1324
+ "grad_norm": 0.2421875,
1325
+ "learning_rate": 0.0002005988023952096,
1326
+ "loss": 0.1239,
1327
+ "step": 1700
1328
+ },
1329
+ {
1330
+ "epoch": 1.8398268398268398,
1331
+ "eval_loss": 0.16243509948253632,
1332
+ "eval_runtime": 54.1441,
1333
+ "eval_samples_per_second": 9.235,
1334
+ "eval_steps_per_second": 0.074,
1335
+ "step": 1700
1336
+ },
1337
+ {
1338
+ "epoch": 1.8506493506493507,
1339
+ "grad_norm": 0.1455078125,
1340
+ "learning_rate": 0.00019872754491017964,
1341
+ "loss": 0.1257,
1342
+ "step": 1710
1343
+ },
1344
+ {
1345
+ "epoch": 1.8614718614718615,
1346
+ "grad_norm": 0.1982421875,
1347
+ "learning_rate": 0.0001968562874251497,
1348
+ "loss": 0.151,
1349
+ "step": 1720
1350
+ },
1351
+ {
1352
+ "epoch": 1.8722943722943723,
1353
+ "grad_norm": 0.30078125,
1354
+ "learning_rate": 0.00019498502994011978,
1355
+ "loss": 0.1238,
1356
+ "step": 1730
1357
+ },
1358
+ {
1359
+ "epoch": 1.883116883116883,
1360
+ "grad_norm": 0.2734375,
1361
+ "learning_rate": 0.00019311377245508982,
1362
+ "loss": 0.1447,
1363
+ "step": 1740
1364
+ },
1365
+ {
1366
+ "epoch": 1.893939393939394,
1367
+ "grad_norm": 0.1728515625,
1368
+ "learning_rate": 0.00019124251497005986,
1369
+ "loss": 0.14,
1370
+ "step": 1750
1371
+ },
1372
+ {
1373
+ "epoch": 1.9047619047619047,
1374
+ "grad_norm": 0.146484375,
1375
+ "learning_rate": 0.00018937125748502996,
1376
+ "loss": 0.118,
1377
+ "step": 1760
1378
+ },
1379
+ {
1380
+ "epoch": 1.9155844155844157,
1381
+ "grad_norm": 0.275390625,
1382
+ "learning_rate": 0.0001875,
1383
+ "loss": 0.126,
1384
+ "step": 1770
1385
+ },
1386
+ {
1387
+ "epoch": 1.9264069264069263,
1388
+ "grad_norm": 0.19140625,
1389
+ "learning_rate": 0.00018562874251497005,
1390
+ "loss": 0.1399,
1391
+ "step": 1780
1392
+ },
1393
+ {
1394
+ "epoch": 1.9372294372294372,
1395
+ "grad_norm": 0.240234375,
1396
+ "learning_rate": 0.00018375748502994014,
1397
+ "loss": 0.1291,
1398
+ "step": 1790
1399
+ },
1400
+ {
1401
+ "epoch": 1.948051948051948,
1402
+ "grad_norm": 0.205078125,
1403
+ "learning_rate": 0.0001818862275449102,
1404
+ "loss": 0.1339,
1405
+ "step": 1800
1406
+ },
1407
+ {
1408
+ "epoch": 1.948051948051948,
1409
+ "eval_loss": 0.1591983437538147,
1410
+ "eval_runtime": 47.2348,
1411
+ "eval_samples_per_second": 10.585,
1412
+ "eval_steps_per_second": 0.085,
1413
+ "step": 1800
1414
+ },
1415
+ {
1416
+ "epoch": 1.9588744588744589,
1417
+ "grad_norm": 0.22265625,
1418
+ "learning_rate": 0.00018001497005988023,
1419
+ "loss": 0.1414,
1420
+ "step": 1810
1421
+ },
1422
+ {
1423
+ "epoch": 1.9696969696969697,
1424
+ "grad_norm": 0.2080078125,
1425
+ "learning_rate": 0.0001781437125748503,
1426
+ "loss": 0.116,
1427
+ "step": 1820
1428
+ },
1429
+ {
1430
+ "epoch": 1.9805194805194806,
1431
+ "grad_norm": 0.265625,
1432
+ "learning_rate": 0.00017627245508982037,
1433
+ "loss": 0.1279,
1434
+ "step": 1830
1435
+ },
1436
+ {
1437
+ "epoch": 1.9913419913419914,
1438
+ "grad_norm": 0.2236328125,
1439
+ "learning_rate": 0.0001744011976047904,
1440
+ "loss": 0.1136,
1441
+ "step": 1840
1442
+ },
1443
+ {
1444
+ "epoch": 2.002164502164502,
1445
+ "grad_norm": 0.11669921875,
1446
+ "learning_rate": 0.00017252994011976048,
1447
+ "loss": 0.1158,
1448
+ "step": 1850
1449
+ },
1450
+ {
1451
+ "epoch": 2.012987012987013,
1452
+ "grad_norm": 0.12353515625,
1453
+ "learning_rate": 0.00017065868263473055,
1454
+ "loss": 0.113,
1455
+ "step": 1860
1456
+ },
1457
+ {
1458
+ "epoch": 2.0238095238095237,
1459
+ "grad_norm": 0.234375,
1460
+ "learning_rate": 0.0001687874251497006,
1461
+ "loss": 0.1127,
1462
+ "step": 1870
1463
+ },
1464
+ {
1465
+ "epoch": 2.034632034632035,
1466
+ "grad_norm": 0.1728515625,
1467
+ "learning_rate": 0.00016691616766467067,
1468
+ "loss": 0.1289,
1469
+ "step": 1880
1470
+ },
1471
+ {
1472
+ "epoch": 2.0454545454545454,
1473
+ "grad_norm": 0.10791015625,
1474
+ "learning_rate": 0.0001650449101796407,
1475
+ "loss": 0.1236,
1476
+ "step": 1890
1477
+ },
1478
+ {
1479
+ "epoch": 2.0562770562770565,
1480
+ "grad_norm": 0.2080078125,
1481
+ "learning_rate": 0.00016317365269461078,
1482
+ "loss": 0.114,
1483
+ "step": 1900
1484
+ },
1485
+ {
1486
+ "epoch": 2.0562770562770565,
1487
+ "eval_loss": 0.15889017283916473,
1488
+ "eval_runtime": 49.4787,
1489
+ "eval_samples_per_second": 10.105,
1490
+ "eval_steps_per_second": 0.081,
1491
+ "step": 1900
1492
+ },
1493
+ {
1494
+ "epoch": 2.067099567099567,
1495
+ "grad_norm": 0.1875,
1496
+ "learning_rate": 0.00016130239520958085,
1497
+ "loss": 0.1202,
1498
+ "step": 1910
1499
+ },
1500
+ {
1501
+ "epoch": 2.0779220779220777,
1502
+ "grad_norm": 0.30078125,
1503
+ "learning_rate": 0.0001594311377245509,
1504
+ "loss": 0.1236,
1505
+ "step": 1920
1506
+ },
1507
+ {
1508
+ "epoch": 2.088744588744589,
1509
+ "grad_norm": 0.1630859375,
1510
+ "learning_rate": 0.00015755988023952096,
1511
+ "loss": 0.1151,
1512
+ "step": 1930
1513
+ },
1514
+ {
1515
+ "epoch": 2.0995670995670994,
1516
+ "grad_norm": 0.1806640625,
1517
+ "learning_rate": 0.00015568862275449103,
1518
+ "loss": 0.116,
1519
+ "step": 1940
1520
+ },
1521
+ {
1522
+ "epoch": 2.1103896103896105,
1523
+ "grad_norm": 0.20703125,
1524
+ "learning_rate": 0.00015381736526946108,
1525
+ "loss": 0.1376,
1526
+ "step": 1950
1527
+ },
1528
+ {
1529
+ "epoch": 2.121212121212121,
1530
+ "grad_norm": 0.1865234375,
1531
+ "learning_rate": 0.00015194610778443112,
1532
+ "loss": 0.1189,
1533
+ "step": 1960
1534
+ },
1535
+ {
1536
+ "epoch": 2.132034632034632,
1537
+ "grad_norm": 0.1103515625,
1538
+ "learning_rate": 0.00015007485029940122,
1539
+ "loss": 0.1122,
1540
+ "step": 1970
1541
+ },
1542
+ {
1543
+ "epoch": 2.142857142857143,
1544
+ "grad_norm": 0.1494140625,
1545
+ "learning_rate": 0.00014820359281437126,
1546
+ "loss": 0.1159,
1547
+ "step": 1980
1548
+ },
1549
+ {
1550
+ "epoch": 2.153679653679654,
1551
+ "grad_norm": 0.1826171875,
1552
+ "learning_rate": 0.00014633233532934133,
1553
+ "loss": 0.1139,
1554
+ "step": 1990
1555
+ },
1556
+ {
1557
+ "epoch": 2.1645021645021645,
1558
+ "grad_norm": 0.234375,
1559
+ "learning_rate": 0.00014446107784431137,
1560
+ "loss": 0.1173,
1561
+ "step": 2000
1562
+ },
1563
+ {
1564
+ "epoch": 2.1645021645021645,
1565
+ "eval_loss": 0.16071972250938416,
1566
+ "eval_runtime": 53.7032,
1567
+ "eval_samples_per_second": 9.31,
1568
+ "eval_steps_per_second": 0.074,
1569
+ "step": 2000
1570
+ },
1571
+ {
1572
+ "epoch": 2.175324675324675,
1573
+ "grad_norm": 0.10498046875,
1574
+ "learning_rate": 0.00014258982035928144,
1575
+ "loss": 0.0994,
1576
+ "step": 2010
1577
+ },
1578
+ {
1579
+ "epoch": 2.186147186147186,
1580
+ "grad_norm": 0.1630859375,
1581
+ "learning_rate": 0.0001407185628742515,
1582
+ "loss": 0.1135,
1583
+ "step": 2020
1584
+ },
1585
+ {
1586
+ "epoch": 2.196969696969697,
1587
+ "grad_norm": 0.1884765625,
1588
+ "learning_rate": 0.00013884730538922155,
1589
+ "loss": 0.1186,
1590
+ "step": 2030
1591
+ },
1592
+ {
1593
+ "epoch": 2.207792207792208,
1594
+ "grad_norm": 0.1904296875,
1595
+ "learning_rate": 0.00013697604790419162,
1596
+ "loss": 0.1066,
1597
+ "step": 2040
1598
+ },
1599
+ {
1600
+ "epoch": 2.2186147186147185,
1601
+ "grad_norm": 0.1923828125,
1602
+ "learning_rate": 0.0001351047904191617,
1603
+ "loss": 0.0979,
1604
+ "step": 2050
1605
+ },
1606
+ {
1607
+ "epoch": 2.2294372294372296,
1608
+ "grad_norm": 0.10546875,
1609
+ "learning_rate": 0.00013323353293413174,
1610
+ "loss": 0.1214,
1611
+ "step": 2060
1612
+ },
1613
+ {
1614
+ "epoch": 2.24025974025974,
1615
+ "grad_norm": 0.1259765625,
1616
+ "learning_rate": 0.00013136227544910178,
1617
+ "loss": 0.1086,
1618
+ "step": 2070
1619
+ },
1620
+ {
1621
+ "epoch": 2.2510822510822512,
1622
+ "grad_norm": 0.2060546875,
1623
+ "learning_rate": 0.00012949101796407188,
1624
+ "loss": 0.0932,
1625
+ "step": 2080
1626
+ },
1627
+ {
1628
+ "epoch": 2.261904761904762,
1629
+ "grad_norm": 0.1318359375,
1630
+ "learning_rate": 0.00012761976047904192,
1631
+ "loss": 0.1053,
1632
+ "step": 2090
1633
+ },
1634
+ {
1635
+ "epoch": 2.2727272727272725,
1636
+ "grad_norm": 0.16015625,
1637
+ "learning_rate": 0.00012574850299401196,
1638
+ "loss": 0.1231,
1639
+ "step": 2100
1640
+ },
1641
+ {
1642
+ "epoch": 2.2727272727272725,
1643
+ "eval_loss": 0.1609112173318863,
1644
+ "eval_runtime": 50.2755,
1645
+ "eval_samples_per_second": 9.945,
1646
+ "eval_steps_per_second": 0.08,
1647
+ "step": 2100
1648
+ },
1649
+ {
1650
+ "epoch": 2.2835497835497836,
1651
+ "grad_norm": 0.1357421875,
1652
+ "learning_rate": 0.00012387724550898203,
1653
+ "loss": 0.1207,
1654
+ "step": 2110
1655
+ },
1656
+ {
1657
+ "epoch": 2.2943722943722946,
1658
+ "grad_norm": 0.16015625,
1659
+ "learning_rate": 0.0001220059880239521,
1660
+ "loss": 0.1202,
1661
+ "step": 2120
1662
+ },
1663
+ {
1664
+ "epoch": 2.3051948051948052,
1665
+ "grad_norm": 0.1943359375,
1666
+ "learning_rate": 0.00012013473053892216,
1667
+ "loss": 0.1215,
1668
+ "step": 2130
1669
+ },
1670
+ {
1671
+ "epoch": 2.316017316017316,
1672
+ "grad_norm": 0.2314453125,
1673
+ "learning_rate": 0.00011826347305389222,
1674
+ "loss": 0.112,
1675
+ "step": 2140
1676
+ },
1677
+ {
1678
+ "epoch": 2.326839826839827,
1679
+ "grad_norm": 0.1767578125,
1680
+ "learning_rate": 0.00011639221556886227,
1681
+ "loss": 0.1233,
1682
+ "step": 2150
1683
+ },
1684
+ {
1685
+ "epoch": 2.3376623376623376,
1686
+ "grad_norm": 0.16015625,
1687
+ "learning_rate": 0.00011452095808383234,
1688
+ "loss": 0.1086,
1689
+ "step": 2160
1690
+ },
1691
+ {
1692
+ "epoch": 2.3484848484848486,
1693
+ "grad_norm": 0.138671875,
1694
+ "learning_rate": 0.0001126497005988024,
1695
+ "loss": 0.1123,
1696
+ "step": 2170
1697
+ },
1698
+ {
1699
+ "epoch": 2.3593073593073592,
1700
+ "grad_norm": 0.193359375,
1701
+ "learning_rate": 0.00011077844311377246,
1702
+ "loss": 0.1145,
1703
+ "step": 2180
1704
+ },
1705
+ {
1706
+ "epoch": 2.3701298701298703,
1707
+ "grad_norm": 0.2216796875,
1708
+ "learning_rate": 0.00010890718562874253,
1709
+ "loss": 0.1117,
1710
+ "step": 2190
1711
+ },
1712
+ {
1713
+ "epoch": 2.380952380952381,
1714
+ "grad_norm": 0.2890625,
1715
+ "learning_rate": 0.00010703592814371257,
1716
+ "loss": 0.12,
1717
+ "step": 2200
1718
+ },
1719
+ {
1720
+ "epoch": 2.380952380952381,
1721
+ "eval_loss": 0.16090266406536102,
1722
+ "eval_runtime": 59.1415,
1723
+ "eval_samples_per_second": 8.454,
1724
+ "eval_steps_per_second": 0.068,
1725
+ "step": 2200
1726
+ },
1727
+ {
1728
+ "epoch": 2.391774891774892,
1729
+ "grad_norm": 0.220703125,
1730
+ "learning_rate": 0.00010516467065868264,
1731
+ "loss": 0.11,
1732
+ "step": 2210
1733
+ },
1734
+ {
1735
+ "epoch": 2.4025974025974026,
1736
+ "grad_norm": 0.1728515625,
1737
+ "learning_rate": 0.0001032934131736527,
1738
+ "loss": 0.1198,
1739
+ "step": 2220
1740
+ },
1741
+ {
1742
+ "epoch": 2.4134199134199132,
1743
+ "grad_norm": 0.23828125,
1744
+ "learning_rate": 0.00010142215568862275,
1745
+ "loss": 0.1179,
1746
+ "step": 2230
1747
+ },
1748
+ {
1749
+ "epoch": 2.4242424242424243,
1750
+ "grad_norm": 0.197265625,
1751
+ "learning_rate": 9.955089820359281e-05,
1752
+ "loss": 0.1146,
1753
+ "step": 2240
1754
+ },
1755
+ {
1756
+ "epoch": 2.435064935064935,
1757
+ "grad_norm": 0.1953125,
1758
+ "learning_rate": 9.767964071856288e-05,
1759
+ "loss": 0.1067,
1760
+ "step": 2250
1761
+ },
1762
+ {
1763
+ "epoch": 2.445887445887446,
1764
+ "grad_norm": 0.158203125,
1765
+ "learning_rate": 9.580838323353294e-05,
1766
+ "loss": 0.105,
1767
+ "step": 2260
1768
+ },
1769
+ {
1770
+ "epoch": 2.4567099567099566,
1771
+ "grad_norm": 0.189453125,
1772
+ "learning_rate": 9.393712574850299e-05,
1773
+ "loss": 0.1222,
1774
+ "step": 2270
1775
+ },
1776
+ {
1777
+ "epoch": 2.4675324675324677,
1778
+ "grad_norm": 0.1376953125,
1779
+ "learning_rate": 9.206586826347306e-05,
1780
+ "loss": 0.1068,
1781
+ "step": 2280
1782
+ },
1783
+ {
1784
+ "epoch": 2.4783549783549783,
1785
+ "grad_norm": 0.1640625,
1786
+ "learning_rate": 9.01946107784431e-05,
1787
+ "loss": 0.1069,
1788
+ "step": 2290
1789
+ },
1790
+ {
1791
+ "epoch": 2.4891774891774894,
1792
+ "grad_norm": 0.1728515625,
1793
+ "learning_rate": 8.832335329341318e-05,
1794
+ "loss": 0.1276,
1795
+ "step": 2300
1796
+ },
1797
+ {
1798
+ "epoch": 2.4891774891774894,
1799
+ "eval_loss": 0.15976013243198395,
1800
+ "eval_runtime": 51.0014,
1801
+ "eval_samples_per_second": 9.804,
1802
+ "eval_steps_per_second": 0.078,
1803
+ "step": 2300
1804
+ },
1805
+ {
1806
+ "epoch": 2.5,
1807
+ "grad_norm": 0.2431640625,
1808
+ "learning_rate": 8.645209580838323e-05,
1809
+ "loss": 0.1248,
1810
+ "step": 2310
1811
+ },
1812
+ {
1813
+ "epoch": 2.5108225108225106,
1814
+ "grad_norm": 0.146484375,
1815
+ "learning_rate": 8.45808383233533e-05,
1816
+ "loss": 0.1062,
1817
+ "step": 2320
1818
+ },
1819
+ {
1820
+ "epoch": 2.5216450216450217,
1821
+ "grad_norm": 0.1279296875,
1822
+ "learning_rate": 8.270958083832336e-05,
1823
+ "loss": 0.0997,
1824
+ "step": 2330
1825
+ },
1826
+ {
1827
+ "epoch": 2.5324675324675323,
1828
+ "grad_norm": 0.1962890625,
1829
+ "learning_rate": 8.083832335329341e-05,
1830
+ "loss": 0.1171,
1831
+ "step": 2340
1832
+ },
1833
+ {
1834
+ "epoch": 2.5432900432900434,
1835
+ "grad_norm": 0.19921875,
1836
+ "learning_rate": 7.896706586826349e-05,
1837
+ "loss": 0.0944,
1838
+ "step": 2350
1839
+ },
1840
+ {
1841
+ "epoch": 2.554112554112554,
1842
+ "grad_norm": 0.22265625,
1843
+ "learning_rate": 7.709580838323353e-05,
1844
+ "loss": 0.1091,
1845
+ "step": 2360
1846
+ },
1847
+ {
1848
+ "epoch": 2.564935064935065,
1849
+ "grad_norm": 0.2099609375,
1850
+ "learning_rate": 7.52245508982036e-05,
1851
+ "loss": 0.1151,
1852
+ "step": 2370
1853
+ },
1854
+ {
1855
+ "epoch": 2.5757575757575757,
1856
+ "grad_norm": 0.1142578125,
1857
+ "learning_rate": 7.335329341317365e-05,
1858
+ "loss": 0.1174,
1859
+ "step": 2380
1860
+ },
1861
+ {
1862
+ "epoch": 2.5865800865800868,
1863
+ "grad_norm": 0.1494140625,
1864
+ "learning_rate": 7.148203592814371e-05,
1865
+ "loss": 0.1219,
1866
+ "step": 2390
1867
+ },
1868
+ {
1869
+ "epoch": 2.5974025974025974,
1870
+ "grad_norm": 0.244140625,
1871
+ "learning_rate": 6.961077844311377e-05,
1872
+ "loss": 0.1251,
1873
+ "step": 2400
1874
+ },
1875
+ {
1876
+ "epoch": 2.5974025974025974,
1877
+ "eval_loss": 0.1569649577140808,
1878
+ "eval_runtime": 50.4125,
1879
+ "eval_samples_per_second": 9.918,
1880
+ "eval_steps_per_second": 0.079,
1881
+ "step": 2400
1882
+ },
1883
+ {
1884
+ "epoch": 2.608225108225108,
1885
+ "grad_norm": 0.34375,
1886
+ "learning_rate": 6.773952095808384e-05,
1887
+ "loss": 0.1171,
1888
+ "step": 2410
1889
+ },
1890
+ {
1891
+ "epoch": 2.619047619047619,
1892
+ "grad_norm": 0.1630859375,
1893
+ "learning_rate": 6.58682634730539e-05,
1894
+ "loss": 0.1114,
1895
+ "step": 2420
1896
+ },
1897
+ {
1898
+ "epoch": 2.62987012987013,
1899
+ "grad_norm": 0.208984375,
1900
+ "learning_rate": 6.399700598802395e-05,
1901
+ "loss": 0.1046,
1902
+ "step": 2430
1903
+ },
1904
+ {
1905
+ "epoch": 2.6406926406926408,
1906
+ "grad_norm": 0.08154296875,
1907
+ "learning_rate": 6.212574850299401e-05,
1908
+ "loss": 0.1081,
1909
+ "step": 2440
1910
+ },
1911
+ {
1912
+ "epoch": 2.6515151515151514,
1913
+ "grad_norm": 0.142578125,
1914
+ "learning_rate": 6.025449101796408e-05,
1915
+ "loss": 0.1042,
1916
+ "step": 2450
1917
+ },
1918
+ {
1919
+ "epoch": 2.6623376623376624,
1920
+ "grad_norm": 0.1474609375,
1921
+ "learning_rate": 5.8383233532934134e-05,
1922
+ "loss": 0.0913,
1923
+ "step": 2460
1924
+ },
1925
+ {
1926
+ "epoch": 2.673160173160173,
1927
+ "grad_norm": 0.330078125,
1928
+ "learning_rate": 5.65119760479042e-05,
1929
+ "loss": 0.1082,
1930
+ "step": 2470
1931
+ },
1932
+ {
1933
+ "epoch": 2.683982683982684,
1934
+ "grad_norm": 0.173828125,
1935
+ "learning_rate": 5.4640718562874254e-05,
1936
+ "loss": 0.1158,
1937
+ "step": 2480
1938
+ },
1939
+ {
1940
+ "epoch": 2.6948051948051948,
1941
+ "grad_norm": 0.1689453125,
1942
+ "learning_rate": 5.276946107784431e-05,
1943
+ "loss": 0.1135,
1944
+ "step": 2490
1945
+ },
1946
+ {
1947
+ "epoch": 2.7056277056277054,
1948
+ "grad_norm": 0.2099609375,
1949
+ "learning_rate": 5.089820359281437e-05,
1950
+ "loss": 0.1332,
1951
+ "step": 2500
1952
+ },
1953
+ {
1954
+ "epoch": 2.7056277056277054,
1955
+ "eval_loss": 0.15841823816299438,
1956
+ "eval_runtime": 48.0255,
1957
+ "eval_samples_per_second": 10.411,
1958
+ "eval_steps_per_second": 0.083,
1959
+ "step": 2500
1960
+ },
1961
+ {
1962
+ "epoch": 2.7164502164502164,
1963
+ "grad_norm": 0.140625,
1964
+ "learning_rate": 4.902694610778443e-05,
1965
+ "loss": 0.1076,
1966
+ "step": 2510
1967
+ },
1968
+ {
1969
+ "epoch": 2.7272727272727275,
1970
+ "grad_norm": 0.11572265625,
1971
+ "learning_rate": 4.7155688622754486e-05,
1972
+ "loss": 0.1089,
1973
+ "step": 2520
1974
+ },
1975
+ {
1976
+ "epoch": 2.738095238095238,
1977
+ "grad_norm": 0.1484375,
1978
+ "learning_rate": 4.5284431137724556e-05,
1979
+ "loss": 0.1061,
1980
+ "step": 2530
1981
+ },
1982
+ {
1983
+ "epoch": 2.7489177489177488,
1984
+ "grad_norm": 0.123046875,
1985
+ "learning_rate": 4.341317365269461e-05,
1986
+ "loss": 0.1067,
1987
+ "step": 2540
1988
+ },
1989
+ {
1990
+ "epoch": 2.75974025974026,
1991
+ "grad_norm": 0.201171875,
1992
+ "learning_rate": 4.154191616766467e-05,
1993
+ "loss": 0.1113,
1994
+ "step": 2550
1995
+ },
1996
+ {
1997
+ "epoch": 2.7705627705627704,
1998
+ "grad_norm": 0.1494140625,
1999
+ "learning_rate": 3.967065868263473e-05,
2000
+ "loss": 0.1056,
2001
+ "step": 2560
2002
+ },
2003
+ {
2004
+ "epoch": 2.7813852813852815,
2005
+ "grad_norm": 0.181640625,
2006
+ "learning_rate": 3.779940119760479e-05,
2007
+ "loss": 0.1135,
2008
+ "step": 2570
2009
+ },
2010
+ {
2011
+ "epoch": 2.792207792207792,
2012
+ "grad_norm": 0.15625,
2013
+ "learning_rate": 3.592814371257485e-05,
2014
+ "loss": 0.1116,
2015
+ "step": 2580
2016
+ },
2017
+ {
2018
+ "epoch": 2.8030303030303028,
2019
+ "grad_norm": 0.2255859375,
2020
+ "learning_rate": 3.405688622754491e-05,
2021
+ "loss": 0.1169,
2022
+ "step": 2590
2023
+ },
2024
+ {
2025
+ "epoch": 2.813852813852814,
2026
+ "grad_norm": 0.08154296875,
2027
+ "learning_rate": 3.2185628742514966e-05,
2028
+ "loss": 0.1023,
2029
+ "step": 2600
2030
+ },
2031
+ {
2032
+ "epoch": 2.813852813852814,
2033
+ "eval_loss": 0.15775570273399353,
2034
+ "eval_runtime": 49.2082,
2035
+ "eval_samples_per_second": 10.161,
2036
+ "eval_steps_per_second": 0.081,
2037
+ "step": 2600
2038
+ },
2039
+ {
2040
+ "epoch": 2.824675324675325,
2041
+ "grad_norm": 0.103515625,
2042
+ "learning_rate": 3.031437125748503e-05,
2043
+ "loss": 0.1076,
2044
+ "step": 2610
2045
+ },
2046
+ {
2047
+ "epoch": 2.8354978354978355,
2048
+ "grad_norm": 0.1611328125,
2049
+ "learning_rate": 2.844311377245509e-05,
2050
+ "loss": 0.1082,
2051
+ "step": 2620
2052
+ },
2053
+ {
2054
+ "epoch": 2.846320346320346,
2055
+ "grad_norm": 0.28515625,
2056
+ "learning_rate": 2.6571856287425152e-05,
2057
+ "loss": 0.1125,
2058
+ "step": 2630
2059
+ },
2060
+ {
2061
+ "epoch": 2.857142857142857,
2062
+ "grad_norm": 0.07568359375,
2063
+ "learning_rate": 2.4700598802395212e-05,
2064
+ "loss": 0.0981,
2065
+ "step": 2640
2066
+ },
2067
+ {
2068
+ "epoch": 2.867965367965368,
2069
+ "grad_norm": 0.169921875,
2070
+ "learning_rate": 2.282934131736527e-05,
2071
+ "loss": 0.1197,
2072
+ "step": 2650
2073
+ },
2074
+ {
2075
+ "epoch": 2.878787878787879,
2076
+ "grad_norm": 0.34765625,
2077
+ "learning_rate": 2.095808383233533e-05,
2078
+ "loss": 0.1214,
2079
+ "step": 2660
2080
+ },
2081
+ {
2082
+ "epoch": 2.8896103896103895,
2083
+ "grad_norm": 0.1513671875,
2084
+ "learning_rate": 1.9086826347305392e-05,
2085
+ "loss": 0.105,
2086
+ "step": 2670
2087
+ },
2088
+ {
2089
+ "epoch": 2.9004329004329006,
2090
+ "grad_norm": 0.1767578125,
2091
+ "learning_rate": 1.7215568862275448e-05,
2092
+ "loss": 0.1101,
2093
+ "step": 2680
2094
+ },
2095
+ {
2096
+ "epoch": 2.911255411255411,
2097
+ "grad_norm": 0.169921875,
2098
+ "learning_rate": 1.5344311377245508e-05,
2099
+ "loss": 0.1283,
2100
+ "step": 2690
2101
+ },
2102
+ {
2103
+ "epoch": 2.9220779220779223,
2104
+ "grad_norm": 0.130859375,
2105
+ "learning_rate": 1.347305389221557e-05,
2106
+ "loss": 0.1126,
2107
+ "step": 2700
2108
+ },
2109
+ {
2110
+ "epoch": 2.9220779220779223,
2111
+ "eval_loss": 0.1572347730398178,
2112
+ "eval_runtime": 58.0947,
2113
+ "eval_samples_per_second": 8.607,
2114
+ "eval_steps_per_second": 0.069,
2115
+ "step": 2700
2116
+ },
2117
+ {
2118
+ "epoch": 2.932900432900433,
2119
+ "grad_norm": 0.197265625,
2120
+ "learning_rate": 1.1601796407185628e-05,
2121
+ "loss": 0.1127,
2122
+ "step": 2710
2123
+ },
2124
+ {
2125
+ "epoch": 2.9437229437229435,
2126
+ "grad_norm": 0.259765625,
2127
+ "learning_rate": 9.73053892215569e-06,
2128
+ "loss": 0.1236,
2129
+ "step": 2720
2130
+ },
2131
+ {
2132
+ "epoch": 2.9545454545454546,
2133
+ "grad_norm": 0.1416015625,
2134
+ "learning_rate": 7.859281437125748e-06,
2135
+ "loss": 0.1137,
2136
+ "step": 2730
2137
+ },
2138
+ {
2139
+ "epoch": 2.965367965367965,
2140
+ "grad_norm": 0.2041015625,
2141
+ "learning_rate": 5.9880239520958085e-06,
2142
+ "loss": 0.1213,
2143
+ "step": 2740
2144
+ },
2145
+ {
2146
+ "epoch": 2.9761904761904763,
2147
+ "grad_norm": 0.1728515625,
2148
+ "learning_rate": 4.116766467065868e-06,
2149
+ "loss": 0.1083,
2150
+ "step": 2750
2151
+ },
2152
+ {
2153
+ "epoch": 2.987012987012987,
2154
+ "grad_norm": 0.2099609375,
2155
+ "learning_rate": 2.2455089820359283e-06,
2156
+ "loss": 0.1217,
2157
+ "step": 2760
2158
+ },
2159
+ {
2160
+ "epoch": 2.997835497835498,
2161
+ "grad_norm": 0.2099609375,
2162
+ "learning_rate": 3.7425149700598803e-07,
2163
+ "loss": 0.1188,
2164
+ "step": 2770
2165
+ }
2166
+ ],
2167
+ "logging_steps": 10,
2168
+ "max_steps": 2772,
2169
+ "num_input_tokens_seen": 0,
2170
+ "num_train_epochs": 3,
2171
+ "save_steps": 500,
2172
+ "stateful_callbacks": {
2173
+ "TrainerControl": {
2174
+ "args": {
2175
+ "should_epoch_stop": false,
2176
+ "should_evaluate": false,
2177
+ "should_log": false,
2178
+ "should_save": true,
2179
+ "should_training_stop": true
2180
+ },
2181
+ "attributes": {}
2182
+ }
2183
+ },
2184
+ "total_flos": 6.887405812854188e+18,
2185
+ "train_batch_size": 128,
2186
+ "trial_name": null,
2187
+ "trial_params": null
2188
+ }
checkpoint-2772/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f2afcf823dba854c82182f2115a254b2688cf06dceefd5def0595b5a4da45c35
3
+ size 5240
git_hash.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ 4585c16aaf819f24684c118bdacfbf1514bb56b9
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
preprocessor_config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_convert_rgb": true,
3
+ "do_image_splitting": true,
4
+ "do_normalize": true,
5
+ "do_pad": true,
6
+ "do_rescale": true,
7
+ "do_resize": true,
8
+ "image_mean": [
9
+ 0.5,
10
+ 0.5,
11
+ 0.5
12
+ ],
13
+ "image_processor_type": "Idefics3ImageProcessor",
14
+ "image_std": [
15
+ 0.5,
16
+ 0.5,
17
+ 0.5
18
+ ],
19
+ "max_image_size": {
20
+ "longest_edge": 384
21
+ },
22
+ "processor_class": "ColIdefics3Processor",
23
+ "resample": 1,
24
+ "rescale_factor": 0.00392156862745098,
25
+ "size": {
26
+ "longest_edge": 1536
27
+ }
28
+ }
results.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"validation_set": {"ndcg_at_1": 0.828, "ndcg_at_3": 0.87764, "ndcg_at_5": 0.88263, "ndcg_at_10": 0.88981, "ndcg_at_20": 0.89434, "ndcg_at_50": 0.89829, "ndcg_at_100": 0.90018, "map_at_1": 0.828, "map_at_3": 0.86633, "map_at_5": 0.86913, "map_at_10": 0.87213, "map_at_20": 0.87336, "map_at_50": 0.87399, "map_at_100": 0.87413, "recall_at_1": 0.828, "recall_at_3": 0.91, "recall_at_5": 0.922, "recall_at_10": 0.944, "recall_at_20": 0.962, "recall_at_50": 0.982, "recall_at_100": 0.994, "precision_at_1": 0.828, "precision_at_3": 0.30333, "precision_at_5": 0.1844, "precision_at_10": 0.0944, "precision_at_20": 0.0481, "precision_at_50": 0.01964, "precision_at_100": 0.00994, "mrr_at_1": 0.824, "mrr_at_3": 0.8636666666666666, "mrr_at_5": 0.8669666666666667, "mrr_at_10": 0.8702642857142857, "mrr_at_20": 0.871532798573975, "mrr_at_50": 0.8722038761027637, "mrr_at_100": 0.872391544636181, "naucs_at_1_max": 0.21773711896722028, "naucs_at_1_std": 0.2788401902005385, "naucs_at_1_diff1": 0.9399880550387059, "naucs_at_3_max": 0.15050316422865537, "naucs_at_3_std": 0.6578586990351711, "naucs_at_3_diff1": 0.9553376906318075, "naucs_at_5_max": 0.1060954296248405, "naucs_at_5_std": 0.7918073212190876, "naucs_at_5_diff1": 0.9656443774090842, "naucs_at_10_max": -0.018807523009205383, "naucs_at_10_std": 0.86072762438309, "naucs_at_10_diff1": 0.9807422969187695, "naucs_at_20_max": -0.3403607056857979, "naucs_at_20_std": 0.9314708339476163, "naucs_at_20_diff1": 1.0, "naucs_at_50_max": -0.5622471210706596, "naucs_at_50_std": 1.0, "naucs_at_50_diff1": 1.0, "naucs_at_100_max": -0.6288515406162857, "naucs_at_100_std": 1.0, "naucs_at_100_diff1": 1.0}, "syntheticDocQA_energy": {"ndcg_at_1": 0.94, "ndcg_at_3": 0.97155, "ndcg_at_5": 0.97155, "ndcg_at_10": 0.97155, "ndcg_at_20": 0.97155, "ndcg_at_50": 0.9735, "ndcg_at_100": 0.9735, "map_at_1": 0.94, "map_at_3": 0.965, "map_at_5": 0.965, "map_at_10": 0.965, "map_at_20": 0.965, "map_at_50": 0.96529, "map_at_100": 0.96529, "recall_at_1": 0.94, "recall_at_3": 0.99, "recall_at_5": 0.99, "recall_at_10": 0.99, "recall_at_20": 0.99, "recall_at_50": 1.0, "recall_at_100": 1.0, "precision_at_1": 0.94, "precision_at_3": 0.33, "precision_at_5": 0.198, "precision_at_10": 0.099, "precision_at_20": 0.0495, "precision_at_50": 0.02, "precision_at_100": 0.01, "mrr_at_1": 0.95, "mrr_at_3": 0.97, "mrr_at_5": 0.97, "mrr_at_10": 0.97, "mrr_at_20": 0.97, "mrr_at_50": 0.9702941176470589, "mrr_at_100": 0.9702941176470589, "naucs_at_1_max": 0.5834111422346712, "naucs_at_1_std": -0.3165266106442549, "naucs_at_1_diff1": 0.9782135076252712, "naucs_at_3_max": 0.7222222222222157, "naucs_at_3_std": -0.5634920634921204, "naucs_at_3_diff1": 1.0, "naucs_at_5_max": 0.7222222222222276, "naucs_at_5_std": -0.5634920634920767, "naucs_at_5_diff1": 1.0, "naucs_at_10_max": 0.7222222222222276, "naucs_at_10_std": -0.5634920634920767, "naucs_at_10_diff1": 1.0, "naucs_at_20_max": 0.7222222222222276, "naucs_at_20_std": -0.5634920634920767, "naucs_at_20_diff1": 1.0, "naucs_at_50_max": NaN, "naucs_at_50_std": NaN, "naucs_at_50_diff1": NaN, "naucs_at_100_max": NaN, "naucs_at_100_std": NaN, "naucs_at_100_diff1": NaN}, "syntheticDocQA_healthcare_industry": {"ndcg_at_1": 0.98, "ndcg_at_3": 0.99262, "ndcg_at_5": 0.99262, "ndcg_at_10": 0.99262, "ndcg_at_20": 0.99262, "ndcg_at_50": 0.99262, "ndcg_at_100": 0.99262, "map_at_1": 0.98, "map_at_3": 0.99, "map_at_5": 0.99, "map_at_10": 0.99, "map_at_20": 0.99, "map_at_50": 0.99, "map_at_100": 0.99, "recall_at_1": 0.98, "recall_at_3": 1.0, "recall_at_5": 1.0, "recall_at_10": 1.0, "recall_at_20": 1.0, "recall_at_50": 1.0, "recall_at_100": 1.0, "precision_at_1": 0.98, "precision_at_3": 0.33333, "precision_at_5": 0.2, "precision_at_10": 0.1, "precision_at_20": 0.05, "precision_at_50": 0.02, "precision_at_100": 0.01, "mrr_at_1": 0.98, "mrr_at_3": 0.99, "mrr_at_5": 0.99, "mrr_at_10": 0.99, "mrr_at_20": 0.99, "mrr_at_50": 0.99, "mrr_at_100": 0.99, "naucs_at_1_max": 0.795751633986929, "naucs_at_1_std": -0.8576097105508799, "naucs_at_1_diff1": 0.9346405228758133, "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_50_max": NaN, "naucs_at_50_std": NaN, "naucs_at_50_diff1": NaN, "naucs_at_100_max": NaN, "naucs_at_100_std": NaN, "naucs_at_100_diff1": NaN}, "syntheticDocQA_artificial_intelligence_test": {"ndcg_at_1": 0.99, "ndcg_at_3": 0.995, "ndcg_at_5": 0.995, "ndcg_at_10": 0.995, "ndcg_at_20": 0.995, "ndcg_at_50": 0.995, "ndcg_at_100": 0.995, "map_at_1": 0.99, "map_at_3": 0.99333, "map_at_5": 0.99333, "map_at_10": 0.99333, "map_at_20": 0.99333, "map_at_50": 0.99333, "map_at_100": 0.99333, "recall_at_1": 0.99, "recall_at_3": 1.0, "recall_at_5": 1.0, "recall_at_10": 1.0, "recall_at_20": 1.0, "recall_at_50": 1.0, "recall_at_100": 1.0, "precision_at_1": 0.99, "precision_at_3": 0.33333, "precision_at_5": 0.2, "precision_at_10": 0.1, "precision_at_20": 0.05, "precision_at_50": 0.02, "precision_at_100": 0.01, "mrr_at_1": 0.99, "mrr_at_3": 0.9933333333333333, "mrr_at_5": 0.9933333333333333, "mrr_at_10": 0.9933333333333333, "mrr_at_20": 0.9933333333333333, "mrr_at_50": 0.9933333333333333, "mrr_at_100": 0.9933333333333333, "naucs_at_1_max": 1.0, "naucs_at_1_std": 0.8692810457516276, "naucs_at_1_diff1": 1.0, "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_50_max": NaN, "naucs_at_50_std": NaN, "naucs_at_50_diff1": NaN, "naucs_at_100_max": NaN, "naucs_at_100_std": NaN, "naucs_at_100_diff1": NaN}, "syntheticDocQA_government_reports": {"ndcg_at_1": 0.93, "ndcg_at_3": 0.97155, "ndcg_at_5": 0.97155, "ndcg_at_10": 0.97155, "ndcg_at_20": 0.97155, "ndcg_at_50": 0.97155, "ndcg_at_100": 0.97155, "map_at_1": 0.93, "map_at_3": 0.96167, "map_at_5": 0.96167, "map_at_10": 0.96167, "map_at_20": 0.96167, "map_at_50": 0.96167, "map_at_100": 0.96167, "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_50": 1.0, "recall_at_100": 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_50": 0.02, "precision_at_100": 0.01, "mrr_at_1": 0.94, "mrr_at_3": 0.9666666666666667, "mrr_at_5": 0.9666666666666667, "mrr_at_10": 0.9666666666666667, "mrr_at_20": 0.9666666666666667, "mrr_at_50": 0.9666666666666667, "mrr_at_100": 0.9666666666666667, "naucs_at_1_max": 0.5822328931572631, "naucs_at_1_std": 0.673536081099108, "naucs_at_1_diff1": 0.8622782446311864, "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_50_max": NaN, "naucs_at_50_std": NaN, "naucs_at_50_diff1": NaN, "naucs_at_100_max": NaN, "naucs_at_100_std": NaN, "naucs_at_100_diff1": NaN}, "infovqa_subsampled": {"ndcg_at_1": 0.87, "ndcg_at_3": 0.89567, "ndcg_at_5": 0.9047, "ndcg_at_10": 0.90993, "ndcg_at_20": 0.91299, "ndcg_at_50": 0.91631, "ndcg_at_100": 0.91664, "map_at_1": 0.87, "map_at_3": 0.88933, "map_at_5": 0.89433, "map_at_10": 0.89652, "map_at_20": 0.89737, "map_at_50": 0.89796, "map_at_100": 0.89799, "recall_at_1": 0.87, "recall_at_3": 0.914, "recall_at_5": 0.936, "recall_at_10": 0.952, "recall_at_20": 0.964, "recall_at_50": 0.98, "recall_at_100": 0.982, "precision_at_1": 0.87, "precision_at_3": 0.30467, "precision_at_5": 0.1872, "precision_at_10": 0.0952, "precision_at_20": 0.0482, "precision_at_50": 0.0196, "precision_at_100": 0.00982, "mrr_at_1": 0.872, "mrr_at_3": 0.8903333333333334, "mrr_at_5": 0.8952333333333333, "mrr_at_10": 0.8973857142857142, "mrr_at_20": 0.8981114902641218, "mrr_at_50": 0.8987658806482948, "mrr_at_100": 0.8987961836785979, "naucs_at_1_max": 0.6437088054506402, "naucs_at_1_std": 0.05869806709620076, "naucs_at_1_diff1": 0.9304969266089019, "naucs_at_3_max": 0.6889344885240949, "naucs_at_3_std": 0.1136842333832744, "naucs_at_3_diff1": 0.9251405988752106, "naucs_at_5_max": 0.7031395891690008, "naucs_at_5_std": 0.26028536414565784, "naucs_at_5_diff1": 0.9203431372549006, "naucs_at_10_max": 0.8295012449424186, "naucs_at_10_std": 0.40411998132586524, "naucs_at_10_diff1": 0.9332788671023946, "naucs_at_20_max": 0.8841944185081392, "naucs_at_20_std": 0.49084448594252705, "naucs_at_20_diff1": 0.9337327523601999, "naucs_at_50_max": 0.9292717086834656, "naucs_at_50_std": 0.8650793650793516, "naucs_at_50_diff1": 0.9477124183006472, "naucs_at_100_max": 0.9214130096482952, "naucs_at_100_std": 0.8646125116713362, "naucs_at_100_diff1": 0.9419026870007192}, "docvqa_subsampled": {"ndcg_at_1": 0.478, "ndcg_at_3": 0.53969, "ndcg_at_5": 0.56094, "ndcg_at_10": 0.57779, "ndcg_at_20": 0.59308, "ndcg_at_50": 0.60319, "ndcg_at_100": 0.60994, "map_at_1": 0.478, "map_at_3": 0.525, "map_at_5": 0.5367, "map_at_10": 0.54367, "map_at_20": 0.54794, "map_at_50": 0.54964, "map_at_100": 0.55022, "recall_at_1": 0.478, "recall_at_3": 0.582, "recall_at_5": 0.634, "recall_at_10": 0.686, "recall_at_20": 0.746, "recall_at_50": 0.796, "recall_at_100": 0.838, "precision_at_1": 0.478, "precision_at_3": 0.194, "precision_at_5": 0.1268, "precision_at_10": 0.0686, "precision_at_20": 0.0373, "precision_at_50": 0.01592, "precision_at_100": 0.00838, "mrr_at_1": 0.482, "mrr_at_3": 0.5266666666666666, "mrr_at_5": 0.5375666666666667, "mrr_at_10": 0.5453944444444444, "mrr_at_20": 0.5492540835755232, "mrr_at_50": 0.5511465806213975, "mrr_at_100": 0.5515877171220239, "naucs_at_1_max": 0.17593079866611316, "naucs_at_1_std": 0.6646824085068347, "naucs_at_1_diff1": 0.8815686129321448, "naucs_at_3_max": 0.08521158919497512, "naucs_at_3_std": 0.7560230713519073, "naucs_at_3_diff1": 0.7969302187589984, "naucs_at_5_max": 0.02873191231399215, "naucs_at_5_std": 0.8071899242551022, "naucs_at_5_diff1": 0.7628796322315227, "naucs_at_10_max": -0.07012897967179643, "naucs_at_10_std": 0.874808909287446, "naucs_at_10_diff1": 0.7255395142360395, "naucs_at_20_max": -0.20666283697498458, "naucs_at_20_std": 0.9340741995875509, "naucs_at_20_diff1": 0.6769916807524927, "naucs_at_50_max": -0.25364250789497006, "naucs_at_50_std": 0.913686581994549, "naucs_at_50_diff1": 0.6194624533926231, "naucs_at_100_max": -0.33117544987926306, "naucs_at_100_std": 0.925380107024548, "naucs_at_100_diff1": 0.5889383698778079}, "arxivqa_subsampled": {"ndcg_at_1": 0.774, "ndcg_at_3": 0.82333, "ndcg_at_5": 0.83168, "ndcg_at_10": 0.84763, "ndcg_at_20": 0.85417, "ndcg_at_50": 0.86064, "ndcg_at_100": 0.86196, "map_at_1": 0.774, "map_at_3": 0.81133, "map_at_5": 0.81603, "map_at_10": 0.82249, "map_at_20": 0.82428, "map_at_50": 0.82536, "map_at_100": 0.82548, "recall_at_1": 0.774, "recall_at_3": 0.858, "recall_at_5": 0.878, "recall_at_10": 0.928, "recall_at_20": 0.954, "recall_at_50": 0.986, "recall_at_100": 0.994, "precision_at_1": 0.774, "precision_at_3": 0.286, "precision_at_5": 0.1756, "precision_at_10": 0.0928, "precision_at_20": 0.0477, "precision_at_50": 0.01972, "precision_at_100": 0.00994, "mrr_at_1": 0.774, "mrr_at_3": 0.8103333333333333, "mrr_at_5": 0.8154333333333332, "mrr_at_10": 0.8216412698412698, "mrr_at_20": 0.8235997861185941, "mrr_at_50": 0.8245970376853287, "mrr_at_100": 0.824717741545969, "naucs_at_1_max": 0.6956030338929966, "naucs_at_1_std": 0.034737257785584987, "naucs_at_1_diff1": 0.9216607432220811, "naucs_at_3_max": 0.7191420090012451, "naucs_at_3_std": 0.10592878346397475, "naucs_at_3_diff1": 0.8826691837320956, "naucs_at_5_max": 0.7499255380853105, "naucs_at_5_std": 0.12009531125080068, "naucs_at_5_diff1": 0.8803514602373368, "naucs_at_10_max": 0.8015224608361871, "naucs_at_10_std": 0.08101203444340487, "naucs_at_10_diff1": 0.8493749351592488, "naucs_at_20_max": 0.7641375390735974, "naucs_at_20_std": 0.1359355336337385, "naucs_at_20_diff1": 0.8352819388625017, "naucs_at_50_max": 0.7675736961451084, "naucs_at_50_std": 0.057823129251703305, "naucs_at_50_diff1": 0.8646125116713221, "naucs_at_100_max": 0.7424525365701481, "naucs_at_100_std": 0.49361967009023777, "naucs_at_100_diff1": 0.8638344226579416}, "tabfquad_subsampled": {"ndcg_at_1": 0.73571, "ndcg_at_3": 0.79375, "ndcg_at_5": 0.80251, "ndcg_at_10": 0.81636, "ndcg_at_20": 0.82817, "ndcg_at_50": 0.83832, "ndcg_at_100": 0.83954, "map_at_1": 0.73571, "map_at_3": 0.78036, "map_at_5": 0.78518, "map_at_10": 0.7909, "map_at_20": 0.79416, "map_at_50": 0.79589, "map_at_100": 0.79601, "recall_at_1": 0.73571, "recall_at_3": 0.83214, "recall_at_5": 0.85357, "recall_at_10": 0.89643, "recall_at_20": 0.94286, "recall_at_50": 0.99286, "recall_at_100": 1.0, "precision_at_1": 0.73571, "precision_at_3": 0.27738, "precision_at_5": 0.17071, "precision_at_10": 0.08964, "precision_at_20": 0.04714, "precision_at_50": 0.01986, "precision_at_100": 0.01, "mrr_at_1": 0.7357142857142858, "mrr_at_3": 0.7797619047619048, "mrr_at_5": 0.7838690476190475, "mrr_at_10": 0.7901204648526078, "mrr_at_20": 0.7934665929308786, "mrr_at_50": 0.7952056267189848, "mrr_at_100": 0.7953287794283443, "naucs_at_1_max": 0.3253261692793548, "naucs_at_1_std": 0.06312608068134669, "naucs_at_1_diff1": 0.8084496777604704, "naucs_at_3_max": 0.3105270885332547, "naucs_at_3_std": 0.07538225978983691, "naucs_at_3_diff1": 0.738448308408667, "naucs_at_5_max": 0.31633840963626375, "naucs_at_5_std": 0.09126654154094667, "naucs_at_5_diff1": 0.7328288707799767, "naucs_at_10_max": 0.1824121090591562, "naucs_at_10_std": -0.010980303373330084, "naucs_at_10_diff1": 0.6897377017367963, "naucs_at_20_max": 0.0169234360410811, "naucs_at_20_std": -0.31098272642390185, "naucs_at_20_diff1": 0.6254668534080327, "naucs_at_50_max": -0.3968253968254018, "naucs_at_50_std": -0.690943043884227, "naucs_at_50_diff1": 0.8611111111110865, "naucs_at_100_max": 1.0, "naucs_at_100_std": 1.0, "naucs_at_100_diff1": 1.0}, "tatdqa": {"ndcg_at_1": 0.66987, "ndcg_at_3": 0.77553, "ndcg_at_5": 0.79538, "ndcg_at_10": 0.80853, "ndcg_at_20": 0.81354, "ndcg_at_50": 0.81796, "ndcg_at_100": 0.81903, "map_at_1": 0.66987, "map_at_3": 0.75005, "map_at_5": 0.76108, "map_at_10": 0.76669, "map_at_20": 0.76814, "map_at_50": 0.76886, "map_at_100": 0.76895, "recall_at_1": 0.66987, "recall_at_3": 0.84907, "recall_at_5": 0.89717, "recall_at_10": 0.93686, "recall_at_20": 0.9561, "recall_at_50": 0.97835, "recall_at_100": 0.98497, "precision_at_1": 0.66987, "precision_at_3": 0.28302, "precision_at_5": 0.17943, "precision_at_10": 0.09369, "precision_at_20": 0.04781, "precision_at_50": 0.01957, "precision_at_100": 0.00985, "mrr_at_1": 0.6650631389055923, "mrr_at_3": 0.7486470234515935, "mrr_at_5": 0.7588995790739627, "mrr_at_10": 0.7649008771678646, "mrr_at_20": 0.766294670223957, "mrr_at_50": 0.7669629679577795, "mrr_at_100": 0.7670583172691432, "naucs_at_1_max": 0.2568730737793943, "naucs_at_1_std": -0.12808142195797126, "naucs_at_1_diff1": 0.8234815744151925, "naucs_at_3_max": 0.32424305760807287, "naucs_at_3_std": 0.0023177179676344157, "naucs_at_3_diff1": 0.7226885160746074, "naucs_at_5_max": 0.3697612224125756, "naucs_at_5_std": 0.11798240403444073, "naucs_at_5_diff1": 0.6833142525048825, "naucs_at_10_max": 0.49587555962938135, "naucs_at_10_std": 0.22528738928970488, "naucs_at_10_diff1": 0.7079879601645019, "naucs_at_20_max": 0.5475720475651525, "naucs_at_20_std": 0.3970102434994993, "naucs_at_20_diff1": 0.7067977425806262, "naucs_at_50_max": 0.6739875444992179, "naucs_at_50_std": 0.5159845368869518, "naucs_at_50_diff1": 0.7293116370600532, "naucs_at_100_max": 0.7557825589788986, "naucs_at_100_std": 0.5991543861889267, "naucs_at_100_diff1": 0.787492165390262}, "shift_project": {"ndcg_at_1": 0.67, "ndcg_at_3": 0.77464, "ndcg_at_5": 0.80778, "ndcg_at_10": 0.81467, "ndcg_at_20": 0.82248, "ndcg_at_50": 0.82621, "ndcg_at_100": 0.82621, "map_at_1": 0.67, "map_at_3": 0.75167, "map_at_5": 0.77017, "map_at_10": 0.77326, "map_at_20": 0.77553, "map_at_50": 0.77603, "map_at_100": 0.77603, "recall_at_1": 0.67, "recall_at_3": 0.84, "recall_at_5": 0.92, "recall_at_10": 0.94, "recall_at_20": 0.97, "recall_at_50": 0.99, "recall_at_100": 0.99, "precision_at_1": 0.67, "precision_at_3": 0.28, "precision_at_5": 0.184, "precision_at_10": 0.094, "precision_at_20": 0.0485, "precision_at_50": 0.0198, "precision_at_100": 0.0099, "mrr_at_1": 0.71, "mrr_at_3": 0.775, "mrr_at_5": 0.7935, "mrr_at_10": 0.7961666666666666, "mrr_at_20": 0.7976639928698752, "mrr_at_50": 0.7981844144772533, "mrr_at_100": 0.7981844144772533, "naucs_at_1_max": 0.24121819505341022, "naucs_at_1_std": -0.1675272130148469, "naucs_at_1_diff1": 0.8373652045843135, "naucs_at_3_max": 0.26165930289641504, "naucs_at_3_std": -0.26902307314678475, "naucs_at_3_diff1": 0.697379725085911, "naucs_at_5_max": -0.003501400560219899, "naucs_at_5_std": -0.6327614379084904, "naucs_at_5_diff1": 0.5134220354808603, "naucs_at_10_max": 0.02956738250855913, "naucs_at_10_std": -0.623171490818547, "naucs_at_10_diff1": 0.5325241207594161, "naucs_at_20_max": -0.20572673513850018, "naucs_at_20_std": -1.1190476190476195, "naucs_at_20_diff1": 0.6150015561780299, "naucs_at_50_max": 0.12278244631185525, "naucs_at_50_std": 0.12278244631185525, "naucs_at_50_diff1": 0.12278244631185525, "naucs_at_100_max": 0.12278244631185525, "naucs_at_100_std": 0.12278244631185525, "naucs_at_100_diff1": 0.12278244631185525}}
special_tokens_map.json ADDED
@@ -0,0 +1,53 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ {
4
+ "content": "<fake_token_around_image>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ {
11
+ "content": "<image>",
12
+ "lstrip": false,
13
+ "normalized": false,
14
+ "rstrip": false,
15
+ "single_word": false
16
+ },
17
+ {
18
+ "content": "<end_of_utterance>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ ],
25
+ "bos_token": {
26
+ "content": "<|im_start|>",
27
+ "lstrip": false,
28
+ "normalized": false,
29
+ "rstrip": false,
30
+ "single_word": false
31
+ },
32
+ "eos_token": {
33
+ "content": "<end_of_utterance>",
34
+ "lstrip": false,
35
+ "normalized": false,
36
+ "rstrip": false,
37
+ "single_word": false
38
+ },
39
+ "pad_token": {
40
+ "content": "<|im_end|>",
41
+ "lstrip": false,
42
+ "normalized": false,
43
+ "rstrip": false,
44
+ "single_word": false
45
+ },
46
+ "unk_token": {
47
+ "content": "<|endoftext|>",
48
+ "lstrip": false,
49
+ "normalized": false,
50
+ "rstrip": false,
51
+ "single_word": false
52
+ }
53
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,182 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "0": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": false,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "1": {
13
+ "content": "<|im_start|>",
14
+ "lstrip": false,
15
+ "normalized": false,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "2": {
21
+ "content": "<|im_end|>",
22
+ "lstrip": false,
23
+ "normalized": false,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ },
28
+ "3": {
29
+ "content": "<repo_name>",
30
+ "lstrip": false,
31
+ "normalized": false,
32
+ "rstrip": false,
33
+ "single_word": false,
34
+ "special": true
35
+ },
36
+ "4": {
37
+ "content": "<reponame>",
38
+ "lstrip": false,
39
+ "normalized": false,
40
+ "rstrip": false,
41
+ "single_word": false,
42
+ "special": true
43
+ },
44
+ "5": {
45
+ "content": "<file_sep>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false,
50
+ "special": true
51
+ },
52
+ "6": {
53
+ "content": "<filename>",
54
+ "lstrip": false,
55
+ "normalized": false,
56
+ "rstrip": false,
57
+ "single_word": false,
58
+ "special": true
59
+ },
60
+ "7": {
61
+ "content": "<gh_stars>",
62
+ "lstrip": false,
63
+ "normalized": false,
64
+ "rstrip": false,
65
+ "single_word": false,
66
+ "special": true
67
+ },
68
+ "8": {
69
+ "content": "<issue_start>",
70
+ "lstrip": false,
71
+ "normalized": false,
72
+ "rstrip": false,
73
+ "single_word": false,
74
+ "special": true
75
+ },
76
+ "9": {
77
+ "content": "<issue_comment>",
78
+ "lstrip": false,
79
+ "normalized": false,
80
+ "rstrip": false,
81
+ "single_word": false,
82
+ "special": true
83
+ },
84
+ "10": {
85
+ "content": "<issue_closed>",
86
+ "lstrip": false,
87
+ "normalized": false,
88
+ "rstrip": false,
89
+ "single_word": false,
90
+ "special": true
91
+ },
92
+ "11": {
93
+ "content": "<jupyter_start>",
94
+ "lstrip": false,
95
+ "normalized": false,
96
+ "rstrip": false,
97
+ "single_word": false,
98
+ "special": true
99
+ },
100
+ "12": {
101
+ "content": "<jupyter_text>",
102
+ "lstrip": false,
103
+ "normalized": false,
104
+ "rstrip": false,
105
+ "single_word": false,
106
+ "special": true
107
+ },
108
+ "13": {
109
+ "content": "<jupyter_code>",
110
+ "lstrip": false,
111
+ "normalized": false,
112
+ "rstrip": false,
113
+ "single_word": false,
114
+ "special": true
115
+ },
116
+ "14": {
117
+ "content": "<jupyter_output>",
118
+ "lstrip": false,
119
+ "normalized": false,
120
+ "rstrip": false,
121
+ "single_word": false,
122
+ "special": true
123
+ },
124
+ "15": {
125
+ "content": "<jupyter_script>",
126
+ "lstrip": false,
127
+ "normalized": false,
128
+ "rstrip": false,
129
+ "single_word": false,
130
+ "special": true
131
+ },
132
+ "16": {
133
+ "content": "<empty_output>",
134
+ "lstrip": false,
135
+ "normalized": false,
136
+ "rstrip": false,
137
+ "single_word": false,
138
+ "special": true
139
+ },
140
+ "49152": {
141
+ "content": "<fake_token_around_image>",
142
+ "lstrip": false,
143
+ "normalized": false,
144
+ "rstrip": false,
145
+ "single_word": false,
146
+ "special": true
147
+ },
148
+ "49153": {
149
+ "content": "<image>",
150
+ "lstrip": false,
151
+ "normalized": false,
152
+ "rstrip": false,
153
+ "single_word": false,
154
+ "special": true
155
+ },
156
+ "49154": {
157
+ "content": "<end_of_utterance>",
158
+ "lstrip": false,
159
+ "normalized": false,
160
+ "rstrip": false,
161
+ "single_word": false,
162
+ "special": true
163
+ }
164
+ },
165
+ "additional_special_tokens": [
166
+ "<fake_token_around_image>",
167
+ "<image>",
168
+ "<end_of_utterance>"
169
+ ],
170
+ "bos_token": "<|im_start|>",
171
+ "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}",
172
+ "clean_up_tokenization_spaces": false,
173
+ "eos_token": "<end_of_utterance>",
174
+ "legacy": false,
175
+ "model_max_length": 16384,
176
+ "pad_token": "<|im_end|>",
177
+ "processor_class": "ColIdefics3Processor",
178
+ "tokenizer_class": "GPT2Tokenizer",
179
+ "truncation_side": "left",
180
+ "unk_token": "<|endoftext|>",
181
+ "vocab_size": 49152
182
+ }
training_config.yml ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ config:
2
+ (): colpali_engine.trainer.colmodel_training.ColModelTrainingConfig
3
+ output_dir: !path ../../../models/colsmolvlm-b64
4
+ processor:
5
+ (): colpali_engine.utils.transformers_wrappers.AllPurposeWrapper
6
+ class_to_instanciate: !ext colpali_engine.models.ColIdefics3Processor
7
+ pretrained_model_name_or_path: "./models/SmolVLM-Instruct" # "./models/paligemma-3b-mix-448"
8
+ # num_image_tokens: 2048
9
+ # max_length: 50
10
+
11
+ model:
12
+ (): colpali_engine.utils.transformers_wrappers.AllPurposeWrapper
13
+ class_to_instanciate: !ext colpali_engine.models.ColIdefics3
14
+ pretrained_model_name_or_path: "./models/SmolVLM-Instruct"
15
+ torch_dtype: !ext torch.bfloat16
16
+ # use_cache: false
17
+ attn_implementation: "flash_attention_2"
18
+ # device_map: "auto"
19
+ # quantization_config:
20
+ # (): transformers.BitsAndBytesConfig
21
+ # load_in_4bit: true
22
+ # bnb_4bit_quant_type: "nf4"
23
+ # bnb_4bit_compute_dtype: "bfloat16"
24
+ # bnb_4bit_use_double_quant: true
25
+
26
+ dataset_loading_func: !ext colpali_engine.utils.dataset_transformation.load_train_set
27
+ eval_dataset_loader: !import ../data/test_data.yaml
28
+
29
+ # max_length: 50
30
+ run_eval: true
31
+ loss_func:
32
+ (): colpali_engine.loss.late_interaction_losses.ColbertPairwiseCELoss
33
+ tr_args:
34
+ (): transformers.training_args.TrainingArguments
35
+ output_dir: null
36
+ overwrite_output_dir: true
37
+ num_train_epochs: 3
38
+ per_device_train_batch_size: 32
39
+ gradient_checkpointing: true
40
+ gradient_checkpointing_kwargs: { "use_reentrant": false }
41
+ # gradient_checkpointing: true
42
+ # 6 x 8 gpus = 48 batch size
43
+ # gradient_accumulation_steps: 4
44
+ per_device_eval_batch_size: 32
45
+ eval_strategy: "steps"
46
+ dataloader_num_workers: 8
47
+ # bf16: true
48
+ save_steps: 500
49
+ logging_steps: 10
50
+ eval_steps: 100
51
+ warmup_steps: 100
52
+ learning_rate: 5e-4
53
+ save_total_limit: 1
54
+ # resume_from_checkpoint: true
55
+ # optim: "paged_adamw_8bit"
56
+ # wandb logging
57
+ # wandb_project: "colqwen2"
58
+ # run_name: "colqwen2-ba32-nolora"
59
+ report_to: "wandb"
60
+
61
+
62
+ peft_config:
63
+ (): peft.LoraConfig
64
+ r: 32
65
+ lora_alpha: 32
66
+ lora_dropout: 0.1
67
+ init_lora_weights: "gaussian"
68
+ bias: "none"
69
+ task_type: "FEATURE_EXTRACTION"
70
+ target_modules: '(.*(model.text_model).*(down_proj|gate_proj|up_proj|k_proj|q_proj|v_proj|o_proj).*$|.*(custom_text_proj).*$)'
71
+ # target_modules: '(.*(language_model).*(down_proj|gate_proj|up_proj|k_proj|q_proj|v_proj|o_proj).*$|.*(custom_text_proj).*$)'
72
+
vocab.json ADDED
The diff for this file is too large to render. See raw diff