avsolatorio
commited on
Commit
•
689ab5c
1
Parent(s):
3ef2b9d
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 +13 -273
- 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 sha256:0261c797ffae126153cbbb1467b83780ff91dc1912396077fc38e284df689975
|
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:a1d631affb6dccd4f3fd4cfc3390be46f868b7efccefa2043b4f74f85561c51c
|
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
@@ -1,298 +1,38 @@
|
|
1 |
{
|
2 |
-
"best_metric": 0.
|
3 |
-
"best_model_checkpoint": "doc-topic-model_eval-00_train-03/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": 0.2465483234714004,
|
13 |
-
"grad_norm": 0.
|
14 |
"learning_rate": 1.9950690335305722e-05,
|
15 |
-
"loss": 0.
|
16 |
"step": 500
|
17 |
},
|
18 |
{
|
19 |
"epoch": 0.4930966469428008,
|
20 |
-
"grad_norm": 0.
|
21 |
"learning_rate": 1.9901380670611442e-05,
|
22 |
-
"loss": 0.
|
23 |
"step": 1000
|
24 |
},
|
25 |
{
|
26 |
"epoch": 0.4930966469428008,
|
27 |
"eval_accuracy": 0.9814660487265615,
|
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": 0.7396449704142012,
|
39 |
-
"grad_norm": 0.39963558316230774,
|
40 |
-
"learning_rate": 1.9852071005917162e-05,
|
41 |
-
"loss": 0.084,
|
42 |
-
"step": 1500
|
43 |
-
},
|
44 |
-
{
|
45 |
-
"epoch": 0.9861932938856016,
|
46 |
-
"grad_norm": 0.37207868695259094,
|
47 |
-
"learning_rate": 1.980276134122288e-05,
|
48 |
-
"loss": 0.0734,
|
49 |
-
"step": 2000
|
50 |
-
},
|
51 |
-
{
|
52 |
-
"epoch": 0.9861932938856016,
|
53 |
-
"eval_accuracy": 0.9814660487265615,
|
54 |
-
"eval_f1": 0.0,
|
55 |
-
"eval_loss": 0.0668005496263504,
|
56 |
-
"eval_precision": 0.0,
|
57 |
-
"eval_recall": 0.0,
|
58 |
-
"eval_runtime": 11.8932,
|
59 |
-
"eval_samples_per_second": 681.903,
|
60 |
-
"eval_steps_per_second": 2.691,
|
61 |
-
"step": 2000
|
62 |
-
},
|
63 |
-
{
|
64 |
-
"epoch": 1.232741617357002,
|
65 |
-
"grad_norm": 0.39202699065208435,
|
66 |
-
"learning_rate": 1.9753451676528602e-05,
|
67 |
-
"loss": 0.0653,
|
68 |
-
"step": 2500
|
69 |
-
},
|
70 |
-
{
|
71 |
-
"epoch": 1.4792899408284024,
|
72 |
-
"grad_norm": 0.3334597647190094,
|
73 |
-
"learning_rate": 1.9704142011834322e-05,
|
74 |
-
"loss": 0.0606,
|
75 |
-
"step": 3000
|
76 |
-
},
|
77 |
-
{
|
78 |
-
"epoch": 1.4792899408284024,
|
79 |
-
"eval_accuracy": 0.9823617784202843,
|
80 |
-
"eval_f1": 0.12648276830209867,
|
81 |
-
"eval_loss": 0.05524122714996338,
|
82 |
-
"eval_precision": 0.7700854700854701,
|
83 |
-
"eval_recall": 0.06889959470826643,
|
84 |
-
"eval_runtime": 14.1698,
|
85 |
-
"eval_samples_per_second": 572.346,
|
86 |
-
"eval_steps_per_second": 2.258,
|
87 |
-
"step": 3000
|
88 |
-
},
|
89 |
-
{
|
90 |
-
"epoch": 1.725838264299803,
|
91 |
-
"grad_norm": 0.3517291843891144,
|
92 |
-
"learning_rate": 1.965483234714004e-05,
|
93 |
-
"loss": 0.055,
|
94 |
-
"step": 3500
|
95 |
-
},
|
96 |
-
{
|
97 |
-
"epoch": 1.972386587771203,
|
98 |
-
"grad_norm": 0.43956857919692993,
|
99 |
-
"learning_rate": 1.9605522682445763e-05,
|
100 |
-
"loss": 0.0532,
|
101 |
-
"step": 4000
|
102 |
-
},
|
103 |
-
{
|
104 |
-
"epoch": 1.972386587771203,
|
105 |
-
"eval_accuracy": 0.9840511926527489,
|
106 |
-
"eval_f1": 0.2943500344892456,
|
107 |
-
"eval_loss": 0.0491451695561409,
|
108 |
-
"eval_precision": 0.8177700348432055,
|
109 |
-
"eval_recall": 0.1794754148505009,
|
110 |
-
"eval_runtime": 14.1385,
|
111 |
-
"eval_samples_per_second": 573.612,
|
112 |
-
"eval_steps_per_second": 2.263,
|
113 |
-
"step": 4000
|
114 |
-
},
|
115 |
-
{
|
116 |
-
"epoch": 2.2189349112426036,
|
117 |
-
"grad_norm": 0.33584991097450256,
|
118 |
-
"learning_rate": 1.9556213017751483e-05,
|
119 |
-
"loss": 0.0483,
|
120 |
-
"step": 4500
|
121 |
-
},
|
122 |
-
{
|
123 |
-
"epoch": 2.465483234714004,
|
124 |
-
"grad_norm": 0.35681256651878357,
|
125 |
-
"learning_rate": 1.95069033530572e-05,
|
126 |
-
"loss": 0.0466,
|
127 |
-
"step": 5000
|
128 |
-
},
|
129 |
-
{
|
130 |
-
"epoch": 2.465483234714004,
|
131 |
-
"eval_accuracy": 0.9850886517283898,
|
132 |
-
"eval_f1": 0.43230993363189985,
|
133 |
-
"eval_loss": 0.04674109071493149,
|
134 |
-
"eval_precision": 0.7342375366568915,
|
135 |
-
"eval_recall": 0.3063393744742678,
|
136 |
-
"eval_runtime": 14.36,
|
137 |
-
"eval_samples_per_second": 564.763,
|
138 |
-
"eval_steps_per_second": 2.228,
|
139 |
-
"step": 5000
|
140 |
-
},
|
141 |
-
{
|
142 |
-
"epoch": 2.712031558185404,
|
143 |
-
"grad_norm": 0.36567422747612,
|
144 |
-
"learning_rate": 1.9457593688362923e-05,
|
145 |
-
"loss": 0.0449,
|
146 |
-
"step": 5500
|
147 |
-
},
|
148 |
-
{
|
149 |
-
"epoch": 2.9585798816568047,
|
150 |
-
"grad_norm": 0.4769591689109802,
|
151 |
-
"learning_rate": 1.940828402366864e-05,
|
152 |
-
"loss": 0.0433,
|
153 |
-
"step": 6000
|
154 |
-
},
|
155 |
-
{
|
156 |
-
"epoch": 2.9585798816568047,
|
157 |
-
"eval_accuracy": 0.9859475317828139,
|
158 |
-
"eval_f1": 0.484694142716075,
|
159 |
-
"eval_loss": 0.042755745351314545,
|
160 |
-
"eval_precision": 0.7564892926670993,
|
161 |
-
"eval_recall": 0.3565802554102623,
|
162 |
-
"eval_runtime": 14.1095,
|
163 |
-
"eval_samples_per_second": 574.791,
|
164 |
-
"eval_steps_per_second": 2.268,
|
165 |
-
"step": 6000
|
166 |
-
},
|
167 |
-
{
|
168 |
-
"epoch": 3.2051282051282053,
|
169 |
-
"grad_norm": 0.4941834807395935,
|
170 |
-
"learning_rate": 1.935897435897436e-05,
|
171 |
-
"loss": 0.0405,
|
172 |
-
"step": 6500
|
173 |
-
},
|
174 |
-
{
|
175 |
-
"epoch": 3.4516765285996054,
|
176 |
-
"grad_norm": 0.39021360874176025,
|
177 |
-
"learning_rate": 1.930966469428008e-05,
|
178 |
-
"loss": 0.0391,
|
179 |
-
"step": 7000
|
180 |
-
},
|
181 |
-
{
|
182 |
-
"epoch": 3.4516765285996054,
|
183 |
-
"eval_accuracy": 0.9866122425840101,
|
184 |
-
"eval_f1": 0.5389046177877574,
|
185 |
-
"eval_loss": 0.04075852409005165,
|
186 |
-
"eval_precision": 0.7450398164394655,
|
187 |
-
"eval_recall": 0.4221151640284469,
|
188 |
-
"eval_runtime": 14.2258,
|
189 |
-
"eval_samples_per_second": 570.09,
|
190 |
-
"eval_steps_per_second": 2.249,
|
191 |
-
"step": 7000
|
192 |
-
},
|
193 |
-
{
|
194 |
-
"epoch": 3.698224852071006,
|
195 |
-
"grad_norm": 0.45788881182670593,
|
196 |
-
"learning_rate": 1.92603550295858e-05,
|
197 |
-
"loss": 0.0381,
|
198 |
-
"step": 7500
|
199 |
-
},
|
200 |
-
{
|
201 |
-
"epoch": 3.9447731755424065,
|
202 |
-
"grad_norm": 0.3994843363761902,
|
203 |
-
"learning_rate": 1.921104536489152e-05,
|
204 |
-
"loss": 0.0378,
|
205 |
-
"step": 8000
|
206 |
-
},
|
207 |
-
{
|
208 |
-
"epoch": 3.9447731755424065,
|
209 |
-
"eval_accuracy": 0.9867312952648214,
|
210 |
-
"eval_f1": 0.5526994744386049,
|
211 |
-
"eval_loss": 0.03952678292989731,
|
212 |
-
"eval_precision": 0.7365338087355151,
|
213 |
-
"eval_recall": 0.4423032805689378,
|
214 |
-
"eval_runtime": 14.2307,
|
215 |
-
"eval_samples_per_second": 569.894,
|
216 |
-
"eval_steps_per_second": 2.249,
|
217 |
-
"step": 8000
|
218 |
-
},
|
219 |
-
{
|
220 |
-
"epoch": 4.191321499013807,
|
221 |
-
"grad_norm": 0.40335726737976074,
|
222 |
-
"learning_rate": 1.916173570019724e-05,
|
223 |
-
"loss": 0.0349,
|
224 |
-
"step": 8500
|
225 |
-
},
|
226 |
-
{
|
227 |
-
"epoch": 4.437869822485207,
|
228 |
-
"grad_norm": 0.43202295899391174,
|
229 |
-
"learning_rate": 1.911242603550296e-05,
|
230 |
-
"loss": 0.0338,
|
231 |
-
"step": 9000
|
232 |
-
},
|
233 |
-
{
|
234 |
-
"epoch": 4.437869822485207,
|
235 |
-
"eval_accuracy": 0.986986408152274,
|
236 |
-
"eval_f1": 0.5843744341843201,
|
237 |
-
"eval_loss": 0.03873027116060257,
|
238 |
-
"eval_precision": 0.7160288408208542,
|
239 |
-
"eval_recall": 0.49361474344268563,
|
240 |
-
"eval_runtime": 14.0731,
|
241 |
-
"eval_samples_per_second": 576.277,
|
242 |
-
"eval_steps_per_second": 2.274,
|
243 |
-
"step": 9000
|
244 |
-
},
|
245 |
-
{
|
246 |
-
"epoch": 4.684418145956608,
|
247 |
-
"grad_norm": 0.3353199064731598,
|
248 |
-
"learning_rate": 1.906311637080868e-05,
|
249 |
-
"loss": 0.034,
|
250 |
-
"step": 9500
|
251 |
-
},
|
252 |
-
{
|
253 |
-
"epoch": 4.930966469428008,
|
254 |
-
"grad_norm": 0.3929939270019531,
|
255 |
-
"learning_rate": 1.90138067061144e-05,
|
256 |
-
"loss": 0.0333,
|
257 |
-
"step": 10000
|
258 |
-
},
|
259 |
-
{
|
260 |
-
"epoch": 4.930966469428008,
|
261 |
-
"eval_accuracy": 0.9870771149567017,
|
262 |
-
"eval_f1": 0.5952951620062139,
|
263 |
-
"eval_loss": 0.037975214421749115,
|
264 |
-
"eval_precision": 0.7094044218766529,
|
265 |
-
"eval_recall": 0.5128087481838343,
|
266 |
-
"eval_runtime": 14.136,
|
267 |
-
"eval_samples_per_second": 573.711,
|
268 |
-
"eval_steps_per_second": 2.264,
|
269 |
-
"step": 10000
|
270 |
-
},
|
271 |
-
{
|
272 |
-
"epoch": 5.177514792899408,
|
273 |
-
"grad_norm": 0.2832840085029602,
|
274 |
-
"learning_rate": 1.896459566074951e-05,
|
275 |
-
"loss": 0.0303,
|
276 |
-
"step": 10500
|
277 |
-
},
|
278 |
-
{
|
279 |
-
"epoch": 5.424063116370808,
|
280 |
-
"grad_norm": 0.43404069542884827,
|
281 |
-
"learning_rate": 1.891528599605523e-05,
|
282 |
-
"loss": 0.0301,
|
283 |
-
"step": 11000
|
284 |
-
},
|
285 |
-
{
|
286 |
-
"epoch": 5.424063116370808,
|
287 |
-
"eval_accuracy": 0.9875660813243193,
|
288 |
-
"eval_f1": 0.6041958041958042,
|
289 |
-
"eval_loss": 0.03708931431174278,
|
290 |
-
"eval_precision": 0.7367957746478874,
|
291 |
-
"eval_recall": 0.5120440467997247,
|
292 |
-
"eval_runtime": 14.265,
|
293 |
-
"eval_samples_per_second": 568.525,
|
294 |
-
"eval_steps_per_second": 2.243,
|
295 |
-
"step": 11000
|
296 |
}
|
297 |
],
|
298 |
"logging_steps": 500,
|
@@ -321,7 +61,7 @@
|
|
321 |
"attributes": {}
|
322 |
}
|
323 |
},
|
324 |
-
"total_flos":
|
325 |
"train_batch_size": 4,
|
326 |
"trial_name": null,
|
327 |
"trial_params": null
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.08951831609010696,
|
3 |
+
"best_model_checkpoint": "doc-topic-model_eval-00_train-03/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": 0.3881610631942749,
|
14 |
"learning_rate": 1.9950690335305722e-05,
|
15 |
+
"loss": 0.1669,
|
16 |
"step": 500
|
17 |
},
|
18 |
{
|
19 |
"epoch": 0.4930966469428008,
|
20 |
+
"grad_norm": 0.3564796447753906,
|
21 |
"learning_rate": 1.9901380670611442e-05,
|
22 |
+
"loss": 0.0935,
|
23 |
"step": 1000
|
24 |
},
|
25 |
{
|
26 |
"epoch": 0.4930966469428008,
|
27 |
"eval_accuracy": 0.9814660487265615,
|
28 |
"eval_f1": 0.0,
|
29 |
+
"eval_loss": 0.08951831609010696,
|
30 |
"eval_precision": 0.0,
|
31 |
"eval_recall": 0.0,
|
32 |
+
"eval_runtime": 12.2339,
|
33 |
+
"eval_samples_per_second": 662.913,
|
34 |
+
"eval_steps_per_second": 2.616,
|
35 |
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
36 |
}
|
37 |
],
|
38 |
"logging_steps": 500,
|
|
|
61 |
"attributes": {}
|
62 |
}
|
63 |
},
|
64 |
+
"total_flos": 22523300934480.0,
|
65 |
"train_batch_size": 4,
|
66 |
"trial_name": null,
|
67 |
"trial_params": null
|
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:b0815310043095bb92c789617e2f93d478e30d81c4c444abbe71b4d0c0c4fde6
|
3 |
size 5240
|