Training in progress, step 1000, checkpoint
Browse files- last-checkpoint/config.json +174 -174
- last-checkpoint/model.safetensors +1 -1
- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/rng_state.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +30 -333
- last-checkpoint/training_args.bin +1 -1
last-checkpoint/config.json
CHANGED
|
@@ -8,184 +8,184 @@
|
|
| 8 |
"hidden_dropout_prob": 0.1,
|
| 9 |
"hidden_size": 768,
|
| 10 |
"id2label": {
|
| 11 |
-
"0": "
|
| 12 |
-
"1": "
|
| 13 |
-
"2": "
|
| 14 |
-
"3": "
|
| 15 |
-
"4": "
|
| 16 |
-
"5": "
|
| 17 |
-
"6": "
|
| 18 |
-
"7": "
|
| 19 |
-
"8": "
|
| 20 |
-
"9": "
|
| 21 |
-
"10": "
|
| 22 |
-
"11": "
|
| 23 |
-
"12": "
|
| 24 |
-
"13": "
|
| 25 |
-
"14": "
|
| 26 |
-
"15": "
|
| 27 |
-
"16": "
|
| 28 |
-
"17": "
|
| 29 |
-
"18": "
|
| 30 |
-
"19": "
|
| 31 |
-
"20": "
|
| 32 |
-
"21": "
|
| 33 |
-
"22": "
|
| 34 |
-
"23": "
|
| 35 |
-
"24": "
|
| 36 |
-
"25": "
|
| 37 |
-
"26": "
|
| 38 |
-
"27": "
|
| 39 |
-
"28": "
|
| 40 |
-
"29": "
|
| 41 |
-
"30": "
|
| 42 |
-
"31": "
|
| 43 |
-
"32": "
|
| 44 |
-
"33": "
|
| 45 |
-
"34": "
|
| 46 |
-
"35": "
|
| 47 |
-
"36": "
|
| 48 |
-
"37": "
|
| 49 |
-
"38": "
|
| 50 |
-
"39": "
|
| 51 |
-
"40": "
|
| 52 |
-
"41": "
|
| 53 |
-
"42": "
|
| 54 |
-
"43": "
|
| 55 |
-
"44": "
|
| 56 |
-
"45": "
|
| 57 |
-
"46": "
|
| 58 |
-
"47": "
|
| 59 |
-
"48": "
|
| 60 |
-
"49": "
|
| 61 |
-
"50": "
|
| 62 |
-
"51": "
|
| 63 |
-
"52": "
|
| 64 |
-
"53": "
|
| 65 |
-
"54": "
|
| 66 |
-
"55": "
|
| 67 |
-
"56": "
|
| 68 |
-
"57": "
|
| 69 |
-
"58": "
|
| 70 |
-
"59": "
|
| 71 |
-
"60": "
|
| 72 |
-
"61": "
|
| 73 |
-
"62": "
|
| 74 |
-
"63": "
|
| 75 |
-
"64": "
|
| 76 |
-
"65": "
|
| 77 |
-
"66": "
|
| 78 |
-
"67": "
|
| 79 |
-
"68": "
|
| 80 |
-
"69": "
|
| 81 |
-
"70": "
|
| 82 |
-
"71": "
|
| 83 |
-
"72": "
|
| 84 |
-
"73": "
|
| 85 |
-
"74": "
|
| 86 |
-
"75": "
|
| 87 |
-
"76": "
|
| 88 |
-
"77": "
|
| 89 |
-
"78": "
|
| 90 |
-
"79": "
|
| 91 |
-
"80": "
|
| 92 |
-
"81": "
|
| 93 |
-
"82": "
|
| 94 |
-
"83": "
|
| 95 |
-
"84": "
|
| 96 |
-
"85": "
|
| 97 |
-
"86": "
|
| 98 |
},
|
| 99 |
"initializer_range": 0.02,
|
| 100 |
"intermediate_size": 3072,
|
| 101 |
"label2id": {
|
| 102 |
-
"Agribusiness and Value Chains":
|
| 103 |
-
"Agriculture and Food":
|
| 104 |
-
"Biodiversity":
|
| 105 |
-
"COVID-19 Hub":
|
| 106 |
-
"Climate Change":
|
| 107 |
-
"Climate-Smart Agriculture":
|
| 108 |
-
"Community-Driven Development":
|
| 109 |
-
"Competition Policy":
|
| 110 |
-
"Competitiveness":
|
| 111 |
-
"Credit Infrastructure":
|
| 112 |
-
"Debt":
|
| 113 |
-
"Debt Relief":
|
| 114 |
-
"Digital Development":
|
| 115 |
-
"Disability Inclusion":
|
| 116 |
-
"Disaster Risk Management":
|
| 117 |
-
"Early Childhood Development":
|
| 118 |
-
"Education":
|
| 119 |
-
"Education and Technology":
|
| 120 |
-
"Energy":
|
| 121 |
-
"Environment":
|
| 122 |
-
"Extractive Industries":
|
| 123 |
-
"Financial Inclusion":
|
| 124 |
-
"Financial Integrity":
|
| 125 |
-
"Financial Sector":
|
| 126 |
-
"Financial Stability":
|
| 127 |
-
"Food Security Update":
|
| 128 |
-
"Food System Jobs":
|
| 129 |
-
"Forests and Landscapes":
|
| 130 |
-
"Fragility, Conflict, and Violence":
|
| 131 |
-
"Gas Flaring Reduction":
|
| 132 |
-
"Gender":
|
| 133 |
-
"Girls' Education":
|
| 134 |
-
"Global Financing Facility for Women, Children and Adolescents":
|
| 135 |
-
"Global Value Chains":
|
| 136 |
-
"Governance":
|
| 137 |
-
"Health":
|
| 138 |
-
"Higher Education":
|
| 139 |
-
"Inclusive Cities":
|
| 140 |
-
"Indigenous Peoples":
|
| 141 |
-
"Inequality and Shared Prosperity":
|
| 142 |
-
"Infectious diseases and Vaccines":
|
| 143 |
-
"Infrastructure":
|
| 144 |
-
"Innovation and Entrepreneurship":
|
| 145 |
-
"Investment Climate":
|
| 146 |
-
"Jobs & Development":
|
| 147 |
-
"Jobs and Development":
|
| 148 |
-
"Labor Markets":
|
| 149 |
-
"Land":
|
| 150 |
-
"Livestock and Sustainability":
|
| 151 |
-
"Long-Term Finance":
|
| 152 |
-
"Macroeconomics":
|
| 153 |
-
"Marine Plastic Pollution":
|
| 154 |
-
"Measuring Poverty":
|
| 155 |
-
"Migration":
|
| 156 |
-
"Mining Investment and Governance Review":
|
| 157 |
-
"Natural Capital":
|
| 158 |
-
"Nutrition":
|
| 159 |
-
"Oceans, Fisheries, and Coastal Economies":
|
| 160 |
-
"One Health":
|
| 161 |
-
"Pandemic Preparedness and COVID-19":
|
| 162 |
-
"Payment Systems":
|
| 163 |
-
"Pensions":
|
| 164 |
-
"Pollution":
|
| 165 |
-
"Poverty":
|
| 166 |
-
"Procurement for Development":
|
| 167 |
-
"Regional Integration":
|
| 168 |
-
"Safety Nets and Cash Transfers":
|
| 169 |
-
"Sanitation":
|
| 170 |
-
"Sexual Orientation and Gender Identity":
|
| 171 |
-
"Skills Development":
|
| 172 |
-
"Small and Medium Enterprises Finance":
|
| 173 |
-
"Social Inclusion":
|
| 174 |
-
"Social Protection":
|
| 175 |
-
"Social Sustainability and Inclusion":
|
| 176 |
-
"Sustainable Communities":
|
| 177 |
-
"Sustainable Infrastructure Finance":
|
| 178 |
-
"Taxes and Government Revenue":
|
| 179 |
-
"Teachers":
|
| 180 |
-
"Trade":
|
| 181 |
-
"Trade Facilitation and Logistics":
|
| 182 |
-
"Transport":
|
| 183 |
-
"Universal Health Coverage":
|
| 184 |
-
"Urban Development":
|
| 185 |
-
"Water":
|
| 186 |
-
"Water Resources Management":
|
| 187 |
-
"Water Supply":
|
| 188 |
-
"Water in Agriculture":
|
| 189 |
},
|
| 190 |
"layer_norm_eps": 1e-07,
|
| 191 |
"max_position_embeddings": 512,
|
|
|
|
| 8 |
"hidden_dropout_prob": 0.1,
|
| 9 |
"hidden_size": 768,
|
| 10 |
"id2label": {
|
| 11 |
+
"0": "Indigenous Peoples",
|
| 12 |
+
"1": "Investment Climate",
|
| 13 |
+
"2": "Marine Plastic Pollution",
|
| 14 |
+
"3": "Poverty",
|
| 15 |
+
"4": "Water",
|
| 16 |
+
"5": "Measuring Poverty",
|
| 17 |
+
"6": "Energy",
|
| 18 |
+
"7": "Sustainable Communities",
|
| 19 |
+
"8": "Disability Inclusion",
|
| 20 |
+
"9": "Macroeconomics",
|
| 21 |
+
"10": "Early Childhood Development",
|
| 22 |
+
"11": "Mining Investment and Governance Review",
|
| 23 |
+
"12": "Infrastructure",
|
| 24 |
+
"13": "Community-Driven Development",
|
| 25 |
+
"14": "Land",
|
| 26 |
+
"15": "Innovation and Entrepreneurship",
|
| 27 |
+
"16": "Competitiveness",
|
| 28 |
+
"17": "Biodiversity",
|
| 29 |
+
"18": "Debt",
|
| 30 |
+
"19": "Higher Education",
|
| 31 |
+
"20": "Livestock and Sustainability",
|
| 32 |
+
"21": "Universal Health Coverage",
|
| 33 |
+
"22": "Nutrition",
|
| 34 |
+
"23": "Education and Technology",
|
| 35 |
+
"24": "Small and Medium Enterprises Finance",
|
| 36 |
+
"25": "Climate-Smart Agriculture",
|
| 37 |
+
"26": "Food Security Update",
|
| 38 |
+
"27": "Inclusive Cities",
|
| 39 |
+
"28": "Jobs & Development",
|
| 40 |
+
"29": "Migration",
|
| 41 |
+
"30": "Infectious diseases and Vaccines",
|
| 42 |
+
"31": "Social Protection",
|
| 43 |
+
"32": "Skills Development",
|
| 44 |
+
"33": "Governance",
|
| 45 |
+
"34": "Pollution",
|
| 46 |
+
"35": "Inequality and Shared Prosperity",
|
| 47 |
+
"36": "Oceans, Fisheries, and Coastal Economies",
|
| 48 |
+
"37": "Global Value Chains",
|
| 49 |
+
"38": "Pensions",
|
| 50 |
+
"39": "Food System Jobs",
|
| 51 |
+
"40": "Debt Relief",
|
| 52 |
+
"41": "Sexual Orientation and Gender Identity",
|
| 53 |
+
"42": "Education",
|
| 54 |
+
"43": "Regional Integration",
|
| 55 |
+
"44": "Jobs and Development",
|
| 56 |
+
"45": "Competition Policy",
|
| 57 |
+
"46": "Fragility, Conflict, and Violence",
|
| 58 |
+
"47": "Urban Development",
|
| 59 |
+
"48": "Pandemic Preparedness and COVID-19",
|
| 60 |
+
"49": "Safety Nets and Cash Transfers",
|
| 61 |
+
"50": "Financial Sector",
|
| 62 |
+
"51": "Climate Change",
|
| 63 |
+
"52": "Transport",
|
| 64 |
+
"53": "COVID-19 Hub",
|
| 65 |
+
"54": "Financial Stability",
|
| 66 |
+
"55": "Gender",
|
| 67 |
+
"56": "Natural Capital",
|
| 68 |
+
"57": "Water in Agriculture",
|
| 69 |
+
"58": "Taxes and Government Revenue",
|
| 70 |
+
"59": "Labor Markets",
|
| 71 |
+
"60": "Trade Facilitation and Logistics",
|
| 72 |
+
"61": "Long-Term Finance",
|
| 73 |
+
"62": "Teachers",
|
| 74 |
+
"63": "Sanitation",
|
| 75 |
+
"64": "Forests and Landscapes",
|
| 76 |
+
"65": "Financial Integrity",
|
| 77 |
+
"66": "Payment Systems",
|
| 78 |
+
"67": "Social Inclusion",
|
| 79 |
+
"68": "Health",
|
| 80 |
+
"69": "Girls' Education",
|
| 81 |
+
"70": "Credit Infrastructure",
|
| 82 |
+
"71": "Social Sustainability and Inclusion",
|
| 83 |
+
"72": "Trade",
|
| 84 |
+
"73": "Financial Inclusion",
|
| 85 |
+
"74": "Global Financing Facility for Women, Children and Adolescents",
|
| 86 |
+
"75": "Sustainable Infrastructure Finance",
|
| 87 |
+
"76": "Extractive Industries",
|
| 88 |
+
"77": "Procurement for Development",
|
| 89 |
+
"78": "Agriculture and Food",
|
| 90 |
+
"79": "Water Supply",
|
| 91 |
+
"80": "Disaster Risk Management",
|
| 92 |
+
"81": "One Health",
|
| 93 |
+
"82": "Environment",
|
| 94 |
+
"83": "Digital Development",
|
| 95 |
+
"84": "Agribusiness and Value Chains",
|
| 96 |
+
"85": "Gas Flaring Reduction",
|
| 97 |
+
"86": "Water Resources Management"
|
| 98 |
},
|
| 99 |
"initializer_range": 0.02,
|
| 100 |
"intermediate_size": 3072,
|
| 101 |
"label2id": {
|
| 102 |
+
"Agribusiness and Value Chains": 84,
|
| 103 |
+
"Agriculture and Food": 78,
|
| 104 |
+
"Biodiversity": 17,
|
| 105 |
+
"COVID-19 Hub": 53,
|
| 106 |
+
"Climate Change": 51,
|
| 107 |
+
"Climate-Smart Agriculture": 25,
|
| 108 |
+
"Community-Driven Development": 13,
|
| 109 |
+
"Competition Policy": 45,
|
| 110 |
+
"Competitiveness": 16,
|
| 111 |
+
"Credit Infrastructure": 70,
|
| 112 |
+
"Debt": 18,
|
| 113 |
+
"Debt Relief": 40,
|
| 114 |
+
"Digital Development": 83,
|
| 115 |
+
"Disability Inclusion": 8,
|
| 116 |
+
"Disaster Risk Management": 80,
|
| 117 |
+
"Early Childhood Development": 10,
|
| 118 |
+
"Education": 42,
|
| 119 |
+
"Education and Technology": 23,
|
| 120 |
+
"Energy": 6,
|
| 121 |
+
"Environment": 82,
|
| 122 |
+
"Extractive Industries": 76,
|
| 123 |
+
"Financial Inclusion": 73,
|
| 124 |
+
"Financial Integrity": 65,
|
| 125 |
+
"Financial Sector": 50,
|
| 126 |
+
"Financial Stability": 54,
|
| 127 |
+
"Food Security Update": 26,
|
| 128 |
+
"Food System Jobs": 39,
|
| 129 |
+
"Forests and Landscapes": 64,
|
| 130 |
+
"Fragility, Conflict, and Violence": 46,
|
| 131 |
+
"Gas Flaring Reduction": 85,
|
| 132 |
+
"Gender": 55,
|
| 133 |
+
"Girls' Education": 69,
|
| 134 |
+
"Global Financing Facility for Women, Children and Adolescents": 74,
|
| 135 |
+
"Global Value Chains": 37,
|
| 136 |
+
"Governance": 33,
|
| 137 |
+
"Health": 68,
|
| 138 |
+
"Higher Education": 19,
|
| 139 |
+
"Inclusive Cities": 27,
|
| 140 |
+
"Indigenous Peoples": 0,
|
| 141 |
+
"Inequality and Shared Prosperity": 35,
|
| 142 |
+
"Infectious diseases and Vaccines": 30,
|
| 143 |
+
"Infrastructure": 12,
|
| 144 |
+
"Innovation and Entrepreneurship": 15,
|
| 145 |
+
"Investment Climate": 1,
|
| 146 |
+
"Jobs & Development": 28,
|
| 147 |
+
"Jobs and Development": 44,
|
| 148 |
+
"Labor Markets": 59,
|
| 149 |
+
"Land": 14,
|
| 150 |
+
"Livestock and Sustainability": 20,
|
| 151 |
+
"Long-Term Finance": 61,
|
| 152 |
+
"Macroeconomics": 9,
|
| 153 |
+
"Marine Plastic Pollution": 2,
|
| 154 |
+
"Measuring Poverty": 5,
|
| 155 |
+
"Migration": 29,
|
| 156 |
+
"Mining Investment and Governance Review": 11,
|
| 157 |
+
"Natural Capital": 56,
|
| 158 |
+
"Nutrition": 22,
|
| 159 |
+
"Oceans, Fisheries, and Coastal Economies": 36,
|
| 160 |
+
"One Health": 81,
|
| 161 |
+
"Pandemic Preparedness and COVID-19": 48,
|
| 162 |
+
"Payment Systems": 66,
|
| 163 |
+
"Pensions": 38,
|
| 164 |
+
"Pollution": 34,
|
| 165 |
+
"Poverty": 3,
|
| 166 |
+
"Procurement for Development": 77,
|
| 167 |
+
"Regional Integration": 43,
|
| 168 |
+
"Safety Nets and Cash Transfers": 49,
|
| 169 |
+
"Sanitation": 63,
|
| 170 |
+
"Sexual Orientation and Gender Identity": 41,
|
| 171 |
+
"Skills Development": 32,
|
| 172 |
+
"Small and Medium Enterprises Finance": 24,
|
| 173 |
+
"Social Inclusion": 67,
|
| 174 |
+
"Social Protection": 31,
|
| 175 |
+
"Social Sustainability and Inclusion": 71,
|
| 176 |
+
"Sustainable Communities": 7,
|
| 177 |
+
"Sustainable Infrastructure Finance": 75,
|
| 178 |
+
"Taxes and Government Revenue": 58,
|
| 179 |
+
"Teachers": 62,
|
| 180 |
+
"Trade": 72,
|
| 181 |
+
"Trade Facilitation and Logistics": 60,
|
| 182 |
+
"Transport": 52,
|
| 183 |
+
"Universal Health Coverage": 21,
|
| 184 |
+
"Urban Development": 47,
|
| 185 |
+
"Water": 4,
|
| 186 |
+
"Water Resources Management": 86,
|
| 187 |
+
"Water Supply": 79,
|
| 188 |
+
"Water in Agriculture": 57
|
| 189 |
},
|
| 190 |
"layer_norm_eps": 1e-07,
|
| 191 |
"max_position_embeddings": 512,
|
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 567860028
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9ce19c672be0d8e93a7fa34f22da21019af73a72b352a65bdfa8a3d06b1c6a9e
|
| 3 |
size 567860028
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1135783354
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a3f1fa73ea7a5820ec872caa2a1dcbe4434a262fc31339d2a61a280dcaae0398
|
| 3 |
size 1135783354
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:34649568145f465443beb487f2295ab3dad9e5f49f758646dd823029413e18fe
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8e136870ce715682b288c29fb4d05a0aec61f1a74b2cc393d7fb9e66e4965261
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,358 +1,55 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "doc-topic-model_eval-01_train-00/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 1000,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch":
|
| 13 |
-
"grad_norm": 0.
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 0.
|
| 16 |
"step": 500
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch":
|
| 20 |
-
"grad_norm": 0.
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss": 0.
|
| 23 |
"step": 1000
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch":
|
| 27 |
"eval_accuracy": 0.981383845685049,
|
| 28 |
"eval_f1": 0.0,
|
| 29 |
-
"eval_loss": 0.
|
| 30 |
"eval_precision": 0.0,
|
| 31 |
"eval_recall": 0.0,
|
| 32 |
-
"eval_runtime":
|
| 33 |
-
"eval_samples_per_second":
|
| 34 |
-
"eval_steps_per_second": 2.
|
| 35 |
"step": 1000
|
| 36 |
-
},
|
| 37 |
-
{
|
| 38 |
-
"epoch": 11.811023622047244,
|
| 39 |
-
"grad_norm": 0.10422395914793015,
|
| 40 |
-
"learning_rate": 4.630905511811024e-06,
|
| 41 |
-
"loss": 0.0907,
|
| 42 |
-
"step": 1500
|
| 43 |
-
},
|
| 44 |
-
{
|
| 45 |
-
"epoch": 15.748031496062993,
|
| 46 |
-
"grad_norm": 0.09992287307977676,
|
| 47 |
-
"learning_rate": 4.507874015748032e-06,
|
| 48 |
-
"loss": 0.0826,
|
| 49 |
-
"step": 2000
|
| 50 |
-
},
|
| 51 |
-
{
|
| 52 |
-
"epoch": 15.748031496062993,
|
| 53 |
-
"eval_accuracy": 0.981383845685049,
|
| 54 |
-
"eval_f1": 0.0,
|
| 55 |
-
"eval_loss": 0.0752759575843811,
|
| 56 |
-
"eval_precision": 0.0,
|
| 57 |
-
"eval_recall": 0.0,
|
| 58 |
-
"eval_runtime": 11.8287,
|
| 59 |
-
"eval_samples_per_second": 685.62,
|
| 60 |
-
"eval_steps_per_second": 2.705,
|
| 61 |
-
"step": 2000
|
| 62 |
-
},
|
| 63 |
-
{
|
| 64 |
-
"epoch": 19.68503937007874,
|
| 65 |
-
"grad_norm": 0.10254054516553879,
|
| 66 |
-
"learning_rate": 4.38484251968504e-06,
|
| 67 |
-
"loss": 0.0738,
|
| 68 |
-
"step": 2500
|
| 69 |
-
},
|
| 70 |
-
{
|
| 71 |
-
"epoch": 23.62204724409449,
|
| 72 |
-
"grad_norm": 0.10703600943088531,
|
| 73 |
-
"learning_rate": 4.261811023622048e-06,
|
| 74 |
-
"loss": 0.0656,
|
| 75 |
-
"step": 3000
|
| 76 |
-
},
|
| 77 |
-
{
|
| 78 |
-
"epoch": 23.62204724409449,
|
| 79 |
-
"eval_accuracy": 0.9818331278257295,
|
| 80 |
-
"eval_f1": 0.0531836312601566,
|
| 81 |
-
"eval_loss": 0.05948049575090408,
|
| 82 |
-
"eval_precision": 0.8933002481389578,
|
| 83 |
-
"eval_recall": 0.027407689379520365,
|
| 84 |
-
"eval_runtime": 14.2215,
|
| 85 |
-
"eval_samples_per_second": 570.263,
|
| 86 |
-
"eval_steps_per_second": 2.25,
|
| 87 |
-
"step": 3000
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 27.559055118110237,
|
| 91 |
-
"grad_norm": 0.10907676815986633,
|
| 92 |
-
"learning_rate": 4.138779527559055e-06,
|
| 93 |
-
"loss": 0.0585,
|
| 94 |
-
"step": 3500
|
| 95 |
-
},
|
| 96 |
-
{
|
| 97 |
-
"epoch": 31.496062992125985,
|
| 98 |
-
"grad_norm": 0.10444527864456177,
|
| 99 |
-
"learning_rate": 4.015748031496064e-06,
|
| 100 |
-
"loss": 0.0529,
|
| 101 |
-
"step": 4000
|
| 102 |
-
},
|
| 103 |
-
{
|
| 104 |
-
"epoch": 31.496062992125985,
|
| 105 |
-
"eval_accuracy": 0.9846790538146463,
|
| 106 |
-
"eval_f1": 0.3410143867349427,
|
| 107 |
-
"eval_loss": 0.049754805862903595,
|
| 108 |
-
"eval_precision": 0.8556133374120526,
|
| 109 |
-
"eval_recall": 0.2129425199847735,
|
| 110 |
-
"eval_runtime": 14.1634,
|
| 111 |
-
"eval_samples_per_second": 572.604,
|
| 112 |
-
"eval_steps_per_second": 2.259,
|
| 113 |
-
"step": 4000
|
| 114 |
-
},
|
| 115 |
-
{
|
| 116 |
-
"epoch": 35.43307086614173,
|
| 117 |
-
"grad_norm": 0.11932671815156937,
|
| 118 |
-
"learning_rate": 3.8927165354330705e-06,
|
| 119 |
-
"loss": 0.0482,
|
| 120 |
-
"step": 4500
|
| 121 |
-
},
|
| 122 |
-
{
|
| 123 |
-
"epoch": 39.37007874015748,
|
| 124 |
-
"grad_norm": 0.11693865060806274,
|
| 125 |
-
"learning_rate": 3.769685039370079e-06,
|
| 126 |
-
"loss": 0.0446,
|
| 127 |
-
"step": 5000
|
| 128 |
-
},
|
| 129 |
-
{
|
| 130 |
-
"epoch": 39.37007874015748,
|
| 131 |
-
"eval_accuracy": 0.9862933514746942,
|
| 132 |
-
"eval_f1": 0.47693222997457946,
|
| 133 |
-
"eval_loss": 0.04449684917926788,
|
| 134 |
-
"eval_precision": 0.8234964512514008,
|
| 135 |
-
"eval_recall": 0.3356680624286258,
|
| 136 |
-
"eval_runtime": 14.0359,
|
| 137 |
-
"eval_samples_per_second": 577.806,
|
| 138 |
-
"eval_steps_per_second": 2.28,
|
| 139 |
-
"step": 5000
|
| 140 |
-
},
|
| 141 |
-
{
|
| 142 |
-
"epoch": 43.30708661417323,
|
| 143 |
-
"grad_norm": 0.12081195414066315,
|
| 144 |
-
"learning_rate": 3.646653543307087e-06,
|
| 145 |
-
"loss": 0.0416,
|
| 146 |
-
"step": 5500
|
| 147 |
-
},
|
| 148 |
-
{
|
| 149 |
-
"epoch": 47.24409448818898,
|
| 150 |
-
"grad_norm": 0.12802477180957794,
|
| 151 |
-
"learning_rate": 3.5236220472440946e-06,
|
| 152 |
-
"loss": 0.039,
|
| 153 |
-
"step": 6000
|
| 154 |
-
},
|
| 155 |
-
{
|
| 156 |
-
"epoch": 47.24409448818898,
|
| 157 |
-
"eval_accuracy": 0.9868857802911122,
|
| 158 |
-
"eval_f1": 0.5325587269512503,
|
| 159 |
-
"eval_loss": 0.041690852493047714,
|
| 160 |
-
"eval_precision": 0.7914414414414415,
|
| 161 |
-
"eval_recall": 0.40129425199847735,
|
| 162 |
-
"eval_runtime": 14.1088,
|
| 163 |
-
"eval_samples_per_second": 574.819,
|
| 164 |
-
"eval_steps_per_second": 2.268,
|
| 165 |
-
"step": 6000
|
| 166 |
-
},
|
| 167 |
-
{
|
| 168 |
-
"epoch": 51.181102362204726,
|
| 169 |
-
"grad_norm": 0.12447524815797806,
|
| 170 |
-
"learning_rate": 3.4005905511811027e-06,
|
| 171 |
-
"loss": 0.0368,
|
| 172 |
-
"step": 6500
|
| 173 |
-
},
|
| 174 |
-
{
|
| 175 |
-
"epoch": 55.118110236220474,
|
| 176 |
-
"grad_norm": 0.14242896437644958,
|
| 177 |
-
"learning_rate": 3.2775590551181103e-06,
|
| 178 |
-
"loss": 0.0348,
|
| 179 |
-
"step": 7000
|
| 180 |
-
},
|
| 181 |
-
{
|
| 182 |
-
"epoch": 55.118110236220474,
|
| 183 |
-
"eval_accuracy": 0.987220261632439,
|
| 184 |
-
"eval_f1": 0.5665112254218547,
|
| 185 |
-
"eval_loss": 0.04018256813287735,
|
| 186 |
-
"eval_precision": 0.7685885729193843,
|
| 187 |
-
"eval_recall": 0.44857251617814997,
|
| 188 |
-
"eval_runtime": 14.0857,
|
| 189 |
-
"eval_samples_per_second": 575.762,
|
| 190 |
-
"eval_steps_per_second": 2.272,
|
| 191 |
-
"step": 7000
|
| 192 |
-
},
|
| 193 |
-
{
|
| 194 |
-
"epoch": 59.05511811023622,
|
| 195 |
-
"grad_norm": 0.12024829536676407,
|
| 196 |
-
"learning_rate": 3.1545275590551183e-06,
|
| 197 |
-
"loss": 0.0332,
|
| 198 |
-
"step": 7500
|
| 199 |
-
},
|
| 200 |
-
{
|
| 201 |
-
"epoch": 62.99212598425197,
|
| 202 |
-
"grad_norm": 0.12477916479110718,
|
| 203 |
-
"learning_rate": 3.0314960629921263e-06,
|
| 204 |
-
"loss": 0.0316,
|
| 205 |
-
"step": 8000
|
| 206 |
-
},
|
| 207 |
-
{
|
| 208 |
-
"epoch": 62.99212598425197,
|
| 209 |
-
"eval_accuracy": 0.9873237240812393,
|
| 210 |
-
"eval_f1": 0.5808023997000374,
|
| 211 |
-
"eval_loss": 0.03934308886528015,
|
| 212 |
-
"eval_precision": 0.7555176198024631,
|
| 213 |
-
"eval_recall": 0.47171678720974497,
|
| 214 |
-
"eval_runtime": 14.1551,
|
| 215 |
-
"eval_samples_per_second": 572.936,
|
| 216 |
-
"eval_steps_per_second": 2.261,
|
| 217 |
-
"step": 8000
|
| 218 |
-
},
|
| 219 |
-
{
|
| 220 |
-
"epoch": 66.92913385826772,
|
| 221 |
-
"grad_norm": 0.13788625597953796,
|
| 222 |
-
"learning_rate": 2.908464566929134e-06,
|
| 223 |
-
"loss": 0.0302,
|
| 224 |
-
"step": 8500
|
| 225 |
-
},
|
| 226 |
-
{
|
| 227 |
-
"epoch": 70.86614173228347,
|
| 228 |
-
"grad_norm": 0.12086493521928787,
|
| 229 |
-
"learning_rate": 2.785433070866142e-06,
|
| 230 |
-
"loss": 0.029,
|
| 231 |
-
"step": 9000
|
| 232 |
-
},
|
| 233 |
-
{
|
| 234 |
-
"epoch": 70.86614173228347,
|
| 235 |
-
"eval_accuracy": 0.9874427767620505,
|
| 236 |
-
"eval_f1": 0.5916297935103245,
|
| 237 |
-
"eval_loss": 0.038920916616916656,
|
| 238 |
-
"eval_precision": 0.7496787758439435,
|
| 239 |
-
"eval_recall": 0.4886181956604492,
|
| 240 |
-
"eval_runtime": 14.1532,
|
| 241 |
-
"eval_samples_per_second": 573.016,
|
| 242 |
-
"eval_steps_per_second": 2.261,
|
| 243 |
-
"step": 9000
|
| 244 |
-
},
|
| 245 |
-
{
|
| 246 |
-
"epoch": 74.80314960629921,
|
| 247 |
-
"grad_norm": 0.1340780407190323,
|
| 248 |
-
"learning_rate": 2.66240157480315e-06,
|
| 249 |
-
"loss": 0.0279,
|
| 250 |
-
"step": 9500
|
| 251 |
-
},
|
| 252 |
-
{
|
| 253 |
-
"epoch": 78.74015748031496,
|
| 254 |
-
"grad_norm": 0.15667644143104553,
|
| 255 |
-
"learning_rate": 2.5393700787401576e-06,
|
| 256 |
-
"loss": 0.0269,
|
| 257 |
-
"step": 10000
|
| 258 |
-
},
|
| 259 |
-
{
|
| 260 |
-
"epoch": 78.74015748031496,
|
| 261 |
-
"eval_accuracy": 0.9875037203962753,
|
| 262 |
-
"eval_f1": 0.598039662639617,
|
| 263 |
-
"eval_loss": 0.03880663216114044,
|
| 264 |
-
"eval_precision": 0.7453409090909091,
|
| 265 |
-
"eval_recall": 0.49935287400076134,
|
| 266 |
-
"eval_runtime": 14.1798,
|
| 267 |
-
"eval_samples_per_second": 571.941,
|
| 268 |
-
"eval_steps_per_second": 2.257,
|
| 269 |
-
"step": 10000
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 82.67716535433071,
|
| 273 |
-
"grad_norm": 0.1377759575843811,
|
| 274 |
-
"learning_rate": 2.4163385826771656e-06,
|
| 275 |
-
"loss": 0.0258,
|
| 276 |
-
"step": 10500
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"epoch": 86.61417322834646,
|
| 280 |
-
"grad_norm": 0.1264483779668808,
|
| 281 |
-
"learning_rate": 2.2933070866141736e-06,
|
| 282 |
-
"loss": 0.0249,
|
| 283 |
-
"step": 11000
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 86.61417322834646,
|
| 287 |
-
"eval_accuracy": 0.9874484459373273,
|
| 288 |
-
"eval_f1": 0.6034745231485628,
|
| 289 |
-
"eval_loss": 0.0388496108353138,
|
| 290 |
-
"eval_precision": 0.7325796282204587,
|
| 291 |
-
"eval_recall": 0.5130567186905215,
|
| 292 |
-
"eval_runtime": 14.1595,
|
| 293 |
-
"eval_samples_per_second": 572.761,
|
| 294 |
-
"eval_steps_per_second": 2.26,
|
| 295 |
-
"step": 11000
|
| 296 |
-
},
|
| 297 |
-
{
|
| 298 |
-
"epoch": 90.55118110236221,
|
| 299 |
-
"grad_norm": 0.1424516886472702,
|
| 300 |
-
"learning_rate": 2.1702755905511812e-06,
|
| 301 |
-
"loss": 0.0241,
|
| 302 |
-
"step": 11500
|
| 303 |
-
},
|
| 304 |
-
{
|
| 305 |
-
"epoch": 94.48818897637796,
|
| 306 |
-
"grad_norm": 0.16246256232261658,
|
| 307 |
-
"learning_rate": 2.0472440944881893e-06,
|
| 308 |
-
"loss": 0.0234,
|
| 309 |
-
"step": 12000
|
| 310 |
-
},
|
| 311 |
-
{
|
| 312 |
-
"epoch": 94.48818897637796,
|
| 313 |
-
"eval_accuracy": 0.9874186827671244,
|
| 314 |
-
"eval_f1": 0.6064985150050978,
|
| 315 |
-
"eval_loss": 0.038960497826337814,
|
| 316 |
-
"eval_precision": 0.7259125636672326,
|
| 317 |
-
"eval_recall": 0.5208222306813856,
|
| 318 |
-
"eval_runtime": 14.1182,
|
| 319 |
-
"eval_samples_per_second": 574.434,
|
| 320 |
-
"eval_steps_per_second": 2.267,
|
| 321 |
-
"step": 12000
|
| 322 |
-
},
|
| 323 |
-
{
|
| 324 |
-
"epoch": 98.4251968503937,
|
| 325 |
-
"grad_norm": 0.15036757290363312,
|
| 326 |
-
"learning_rate": 1.924212598425197e-06,
|
| 327 |
-
"loss": 0.0227,
|
| 328 |
-
"step": 12500
|
| 329 |
-
},
|
| 330 |
-
{
|
| 331 |
-
"epoch": 102.36220472440945,
|
| 332 |
-
"grad_norm": 0.14854416251182556,
|
| 333 |
-
"learning_rate": 1.801181102362205e-06,
|
| 334 |
-
"loss": 0.022,
|
| 335 |
-
"step": 13000
|
| 336 |
-
},
|
| 337 |
-
{
|
| 338 |
-
"epoch": 102.36220472440945,
|
| 339 |
-
"eval_accuracy": 0.987407344416571,
|
| 340 |
-
"eval_f1": 0.6050584522380762,
|
| 341 |
-
"eval_loss": 0.03925173729658127,
|
| 342 |
-
"eval_precision": 0.7269814142277291,
|
| 343 |
-
"eval_recall": 0.5181575942139323,
|
| 344 |
-
"eval_runtime": 14.1509,
|
| 345 |
-
"eval_samples_per_second": 573.11,
|
| 346 |
-
"eval_steps_per_second": 2.261,
|
| 347 |
-
"step": 13000
|
| 348 |
}
|
| 349 |
],
|
| 350 |
"logging_steps": 500,
|
| 351 |
-
"max_steps":
|
| 352 |
"num_input_tokens_seen": 0,
|
| 353 |
-
"num_train_epochs":
|
| 354 |
"save_steps": 1000,
|
| 355 |
"stateful_callbacks": {
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 356 |
"TrainerControl": {
|
| 357 |
"args": {
|
| 358 |
"should_epoch_stop": false,
|
|
@@ -364,8 +61,8 @@
|
|
| 364 |
"attributes": {}
|
| 365 |
}
|
| 366 |
},
|
| 367 |
-
"total_flos":
|
| 368 |
-
"train_batch_size":
|
| 369 |
"trial_name": null,
|
| 370 |
"trial_params": null
|
| 371 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.0910971537232399,
|
| 3 |
+
"best_model_checkpoint": "doc-topic-model_eval-01_train-00/checkpoint-1000",
|
| 4 |
+
"epoch": 0.4930966469428008,
|
| 5 |
"eval_steps": 1000,
|
| 6 |
+
"global_step": 1000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.2465483234714004,
|
| 13 |
+
"grad_norm": 0.32839319109916687,
|
| 14 |
+
"learning_rate": 1.9950690335305722e-05,
|
| 15 |
+
"loss": 0.1668,
|
| 16 |
"step": 500
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.4930966469428008,
|
| 20 |
+
"grad_norm": 0.4088590443134308,
|
| 21 |
+
"learning_rate": 1.9901380670611442e-05,
|
| 22 |
+
"loss": 0.0929,
|
| 23 |
"step": 1000
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.4930966469428008,
|
| 27 |
"eval_accuracy": 0.981383845685049,
|
| 28 |
"eval_f1": 0.0,
|
| 29 |
+
"eval_loss": 0.0910971537232399,
|
| 30 |
"eval_precision": 0.0,
|
| 31 |
"eval_recall": 0.0,
|
| 32 |
+
"eval_runtime": 12.1172,
|
| 33 |
+
"eval_samples_per_second": 669.294,
|
| 34 |
+
"eval_steps_per_second": 2.641,
|
| 35 |
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 36 |
}
|
| 37 |
],
|
| 38 |
"logging_steps": 500,
|
| 39 |
+
"max_steps": 202800,
|
| 40 |
"num_input_tokens_seen": 0,
|
| 41 |
+
"num_train_epochs": 100,
|
| 42 |
"save_steps": 1000,
|
| 43 |
"stateful_callbacks": {
|
| 44 |
+
"EarlyStoppingCallback": {
|
| 45 |
+
"args": {
|
| 46 |
+
"early_stopping_patience": 5,
|
| 47 |
+
"early_stopping_threshold": 0.0
|
| 48 |
+
},
|
| 49 |
+
"attributes": {
|
| 50 |
+
"early_stopping_patience_counter": 0
|
| 51 |
+
}
|
| 52 |
+
},
|
| 53 |
"TrainerControl": {
|
| 54 |
"args": {
|
| 55 |
"should_epoch_stop": false,
|
|
|
|
| 61 |
"attributes": {}
|
| 62 |
}
|
| 63 |
},
|
| 64 |
+
"total_flos": 22568907250224.0,
|
| 65 |
+
"train_batch_size": 4,
|
| 66 |
"trial_name": null,
|
| 67 |
"trial_params": null
|
| 68 |
}
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5240
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:793ce0ea9b102975f2c13ebb162cd3b8fa0b532090e49ab65ff0a4403515a81b
|
| 3 |
size 5240
|