mikhail-panzo
commited on
Commit
•
e914377
1
Parent(s):
f0232ab
Training in progress, step 1000, checkpoint
Browse files
checkpoint-1000/model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 577789320
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0688e482e9f88ba8f20e835e53a1657f34866a68c245056f10f68866513382a3
|
3 |
size 577789320
|
checkpoint-1000/optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1155772233
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:69fec42678b24e36e3673f3d90b5466c0cd5b03dfdf7885abb9076f018576ea8
|
3 |
size 1155772233
|
checkpoint-1000/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 sha256:6782ee4581b95437cf5eb6f04357f25869f5c488a3f0d2d0eee869ccecf25d57
|
3 |
size 14244
|
checkpoint-1000/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 sha256:45070bc91d144eb7f163f351456cac0db81ee5e61490751953caec092c3e6eb7
|
3 |
size 1064
|
checkpoint-1000/trainer_state.json
CHANGED
@@ -1,7 +1,7 @@
|
|
1 |
{
|
2 |
-
"best_metric": 0.
|
3 |
"best_model_checkpoint": "mikhail-panzo/fil_b32_le5_s4000/checkpoint-1000",
|
4 |
-
"epoch":
|
5 |
"eval_steps": 500,
|
6 |
"global_step": 1000,
|
7 |
"is_hyper_param_search": false,
|
@@ -9,166 +9,166 @@
|
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
-
"epoch":
|
13 |
-
"grad_norm": 4.
|
14 |
-
"learning_rate": 2.
|
15 |
-
"loss": 0.
|
16 |
"step": 50
|
17 |
},
|
18 |
{
|
19 |
-
"epoch":
|
20 |
-
"grad_norm":
|
21 |
-
"learning_rate":
|
22 |
-
"loss": 0.
|
23 |
"step": 100
|
24 |
},
|
25 |
{
|
26 |
-
"epoch":
|
27 |
-
"grad_norm":
|
28 |
-
"learning_rate": 7.
|
29 |
-
"loss": 0.
|
30 |
"step": 150
|
31 |
},
|
32 |
{
|
33 |
-
"epoch":
|
34 |
-
"grad_norm": 2.
|
35 |
-
"learning_rate":
|
36 |
-
"loss": 0.
|
37 |
"step": 200
|
38 |
},
|
39 |
{
|
40 |
-
"epoch":
|
41 |
-
"grad_norm":
|
42 |
-
"learning_rate": 1.
|
43 |
-
"loss": 0.
|
44 |
"step": 250
|
45 |
},
|
46 |
{
|
47 |
-
"epoch":
|
48 |
-
"grad_norm": 1.
|
49 |
-
"learning_rate": 1.
|
50 |
-
"loss": 0.
|
51 |
"step": 300
|
52 |
},
|
53 |
{
|
54 |
-
"epoch":
|
55 |
-
"grad_norm":
|
56 |
-
"learning_rate": 1.
|
57 |
-
"loss": 0.
|
58 |
"step": 350
|
59 |
},
|
60 |
{
|
61 |
-
"epoch":
|
62 |
-
"grad_norm":
|
63 |
-
"learning_rate":
|
64 |
-
"loss": 0.
|
65 |
"step": 400
|
66 |
},
|
67 |
{
|
68 |
-
"epoch":
|
69 |
-
"grad_norm":
|
70 |
-
"learning_rate": 2.
|
71 |
-
"loss": 0.
|
72 |
"step": 450
|
73 |
},
|
74 |
{
|
75 |
-
"epoch":
|
76 |
-
"grad_norm": 1.
|
77 |
-
"learning_rate": 2.
|
78 |
-
"loss": 0.
|
79 |
"step": 500
|
80 |
},
|
81 |
{
|
82 |
-
"epoch":
|
83 |
-
"eval_loss": 0.
|
84 |
-
"eval_runtime": 9.
|
85 |
-
"eval_samples_per_second":
|
86 |
-
"eval_steps_per_second": 2.
|
87 |
"step": 500
|
88 |
},
|
89 |
{
|
90 |
-
"epoch":
|
91 |
-
"grad_norm":
|
92 |
-
"learning_rate": 2.
|
93 |
-
"loss": 0.
|
94 |
"step": 550
|
95 |
},
|
96 |
{
|
97 |
-
"epoch":
|
98 |
-
"grad_norm":
|
99 |
-
"learning_rate":
|
100 |
-
"loss": 0.
|
101 |
"step": 600
|
102 |
},
|
103 |
{
|
104 |
-
"epoch":
|
105 |
-
"grad_norm": 1.
|
106 |
-
"learning_rate": 3.
|
107 |
-
"loss": 0.
|
108 |
"step": 650
|
109 |
},
|
110 |
{
|
111 |
-
"epoch":
|
112 |
-
"grad_norm":
|
113 |
-
"learning_rate": 3.
|
114 |
-
"loss": 0.
|
115 |
"step": 700
|
116 |
},
|
117 |
{
|
118 |
-
"epoch":
|
119 |
-
"grad_norm":
|
120 |
-
"learning_rate": 3.
|
121 |
-
"loss": 0.
|
122 |
"step": 750
|
123 |
},
|
124 |
{
|
125 |
-
"epoch":
|
126 |
-
"grad_norm": 1.
|
127 |
-
"learning_rate":
|
128 |
-
"loss": 0.
|
129 |
"step": 800
|
130 |
},
|
131 |
{
|
132 |
-
"epoch":
|
133 |
-
"grad_norm":
|
134 |
-
"learning_rate": 4.
|
135 |
-
"loss": 0.
|
136 |
"step": 850
|
137 |
},
|
138 |
{
|
139 |
-
"epoch":
|
140 |
-
"grad_norm":
|
141 |
-
"learning_rate": 4.
|
142 |
-
"loss": 0.
|
143 |
"step": 900
|
144 |
},
|
145 |
{
|
146 |
-
"epoch":
|
147 |
-
"grad_norm":
|
148 |
-
"learning_rate": 4.
|
149 |
-
"loss": 0.
|
150 |
"step": 950
|
151 |
},
|
152 |
{
|
153 |
-
"epoch":
|
154 |
-
"grad_norm":
|
155 |
-
"learning_rate":
|
156 |
-
"loss": 0.
|
157 |
"step": 1000
|
158 |
},
|
159 |
{
|
160 |
-
"epoch":
|
161 |
-
"eval_loss": 0.
|
162 |
-
"eval_runtime": 9.
|
163 |
-
"eval_samples_per_second": 17.
|
164 |
-
"eval_steps_per_second": 2.
|
165 |
"step": 1000
|
166 |
}
|
167 |
],
|
168 |
"logging_steps": 50,
|
169 |
"max_steps": 4000,
|
170 |
"num_input_tokens_seen": 0,
|
171 |
-
"num_train_epochs":
|
172 |
"save_steps": 500,
|
173 |
"stateful_callbacks": {
|
174 |
"TrainerControl": {
|
@@ -182,7 +182,7 @@
|
|
182 |
"attributes": {}
|
183 |
}
|
184 |
},
|
185 |
-
"total_flos":
|
186 |
"train_batch_size": 16,
|
187 |
"trial_name": null,
|
188 |
"trial_params": null
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.4438903331756592,
|
3 |
"best_model_checkpoint": "mikhail-panzo/fil_b32_le5_s4000/checkpoint-1000",
|
4 |
+
"epoch": 88.88888888888889,
|
5 |
"eval_steps": 500,
|
6 |
"global_step": 1000,
|
7 |
"is_hyper_param_search": false,
|
|
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
+
"epoch": 4.444444444444445,
|
13 |
+
"grad_norm": 4.288165092468262,
|
14 |
+
"learning_rate": 2.5000000000000004e-07,
|
15 |
+
"loss": 0.7989,
|
16 |
"step": 50
|
17 |
},
|
18 |
{
|
19 |
+
"epoch": 8.88888888888889,
|
20 |
+
"grad_norm": 2.4493463039398193,
|
21 |
+
"learning_rate": 5.000000000000001e-07,
|
22 |
+
"loss": 0.7701,
|
23 |
"step": 100
|
24 |
},
|
25 |
{
|
26 |
+
"epoch": 13.333333333333334,
|
27 |
+
"grad_norm": 1.3798907995224,
|
28 |
+
"learning_rate": 7.5e-07,
|
29 |
+
"loss": 0.7374,
|
30 |
"step": 150
|
31 |
},
|
32 |
{
|
33 |
+
"epoch": 17.77777777777778,
|
34 |
+
"grad_norm": 2.6173877716064453,
|
35 |
+
"learning_rate": 1.0000000000000002e-06,
|
36 |
+
"loss": 0.7101,
|
37 |
"step": 200
|
38 |
},
|
39 |
{
|
40 |
+
"epoch": 22.22222222222222,
|
41 |
+
"grad_norm": 1.3756701946258545,
|
42 |
+
"learning_rate": 1.25e-06,
|
43 |
+
"loss": 0.6887,
|
44 |
"step": 250
|
45 |
},
|
46 |
{
|
47 |
+
"epoch": 26.666666666666668,
|
48 |
+
"grad_norm": 1.0456843376159668,
|
49 |
+
"learning_rate": 1.5e-06,
|
50 |
+
"loss": 0.672,
|
51 |
"step": 300
|
52 |
},
|
53 |
{
|
54 |
+
"epoch": 31.11111111111111,
|
55 |
+
"grad_norm": 0.9518398642539978,
|
56 |
+
"learning_rate": 1.75e-06,
|
57 |
+
"loss": 0.6583,
|
58 |
"step": 350
|
59 |
},
|
60 |
{
|
61 |
+
"epoch": 35.55555555555556,
|
62 |
+
"grad_norm": 1.305978775024414,
|
63 |
+
"learning_rate": 2.0000000000000003e-06,
|
64 |
+
"loss": 0.6258,
|
65 |
"step": 400
|
66 |
},
|
67 |
{
|
68 |
+
"epoch": 40.0,
|
69 |
+
"grad_norm": 0.8857623338699341,
|
70 |
+
"learning_rate": 2.25e-06,
|
71 |
+
"loss": 0.5835,
|
72 |
"step": 450
|
73 |
},
|
74 |
{
|
75 |
+
"epoch": 44.44444444444444,
|
76 |
+
"grad_norm": 1.052182912826538,
|
77 |
+
"learning_rate": 2.5e-06,
|
78 |
+
"loss": 0.5539,
|
79 |
"step": 500
|
80 |
},
|
81 |
{
|
82 |
+
"epoch": 44.44444444444444,
|
83 |
+
"eval_loss": 0.48496538400650024,
|
84 |
+
"eval_runtime": 9.4574,
|
85 |
+
"eval_samples_per_second": 16.812,
|
86 |
+
"eval_steps_per_second": 2.115,
|
87 |
"step": 500
|
88 |
},
|
89 |
{
|
90 |
+
"epoch": 48.888888888888886,
|
91 |
+
"grad_norm": 1.249458909034729,
|
92 |
+
"learning_rate": 2.7500000000000004e-06,
|
93 |
+
"loss": 0.5387,
|
94 |
"step": 550
|
95 |
},
|
96 |
{
|
97 |
+
"epoch": 53.333333333333336,
|
98 |
+
"grad_norm": 0.801243245601654,
|
99 |
+
"learning_rate": 3e-06,
|
100 |
+
"loss": 0.5344,
|
101 |
"step": 600
|
102 |
},
|
103 |
{
|
104 |
+
"epoch": 57.77777777777778,
|
105 |
+
"grad_norm": 1.1922223567962646,
|
106 |
+
"learning_rate": 3.2500000000000002e-06,
|
107 |
+
"loss": 0.5259,
|
108 |
"step": 650
|
109 |
},
|
110 |
{
|
111 |
+
"epoch": 62.22222222222222,
|
112 |
+
"grad_norm": 0.9428519606590271,
|
113 |
+
"learning_rate": 3.5e-06,
|
114 |
+
"loss": 0.5179,
|
115 |
"step": 700
|
116 |
},
|
117 |
{
|
118 |
+
"epoch": 66.66666666666667,
|
119 |
+
"grad_norm": 0.755260705947876,
|
120 |
+
"learning_rate": 3.7500000000000005e-06,
|
121 |
+
"loss": 0.5082,
|
122 |
"step": 750
|
123 |
},
|
124 |
{
|
125 |
+
"epoch": 71.11111111111111,
|
126 |
+
"grad_norm": 1.8587590456008911,
|
127 |
+
"learning_rate": 4.000000000000001e-06,
|
128 |
+
"loss": 0.5096,
|
129 |
"step": 800
|
130 |
},
|
131 |
{
|
132 |
+
"epoch": 75.55555555555556,
|
133 |
+
"grad_norm": 0.6854829788208008,
|
134 |
+
"learning_rate": 4.25e-06,
|
135 |
+
"loss": 0.4952,
|
136 |
"step": 850
|
137 |
},
|
138 |
{
|
139 |
+
"epoch": 80.0,
|
140 |
+
"grad_norm": 0.9088854789733887,
|
141 |
+
"learning_rate": 4.5e-06,
|
142 |
+
"loss": 0.4941,
|
143 |
"step": 900
|
144 |
},
|
145 |
{
|
146 |
+
"epoch": 84.44444444444444,
|
147 |
+
"grad_norm": 0.8932746648788452,
|
148 |
+
"learning_rate": 4.75e-06,
|
149 |
+
"loss": 0.4936,
|
150 |
"step": 950
|
151 |
},
|
152 |
{
|
153 |
+
"epoch": 88.88888888888889,
|
154 |
+
"grad_norm": 0.9161014556884766,
|
155 |
+
"learning_rate": 5e-06,
|
156 |
+
"loss": 0.4867,
|
157 |
"step": 1000
|
158 |
},
|
159 |
{
|
160 |
+
"epoch": 88.88888888888889,
|
161 |
+
"eval_loss": 0.4438903331756592,
|
162 |
+
"eval_runtime": 9.2831,
|
163 |
+
"eval_samples_per_second": 17.128,
|
164 |
+
"eval_steps_per_second": 2.154,
|
165 |
"step": 1000
|
166 |
}
|
167 |
],
|
168 |
"logging_steps": 50,
|
169 |
"max_steps": 4000,
|
170 |
"num_input_tokens_seen": 0,
|
171 |
+
"num_train_epochs": 364,
|
172 |
"save_steps": 500,
|
173 |
"stateful_callbacks": {
|
174 |
"TrainerControl": {
|
|
|
182 |
"attributes": {}
|
183 |
}
|
184 |
},
|
185 |
+
"total_flos": 2.752618142335248e+16,
|
186 |
"train_batch_size": 16,
|
187 |
"trial_name": null,
|
188 |
"trial_params": null
|
checkpoint-1000/training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5304
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2de69a241d012253ce4b1526ecd0f9fd0f7dfbedfea74bfff08144d42d9f523a
|
3 |
size 5304
|